This week’s issue features guest expert Ashley Koff, RD, bestselling author of Your Best Shot.

Every Indian kitchen I have been in calls dal “protein.” I have been calling it that for years till I dug in further and looked at the macros.

To hit 30 grams of protein from cooked lentils, you need about 1¾ cups of cooked dal. That serving carries roughly 380 calories and 67 grams of carbohydrate. Just the lentils, no rice or roti.

Sure, the protein is there, but so are the carbs and a high amount of calories.

This issue is about that “everything else.” Vegetarians can get enough protein. It’s just harder than it is on a diet that includes animal-based sources. Plant sources are less protein-dense, and have a higher percentage of carbohydrates or fats. Their amino acid profiles are also less complete, though that matters less than most people think. In this issue, we dig into the details, look at various options, and offer a framework on how to choose between the different options.

TL;DR

• Vegetarian diets can meet protein needs. The plate design is harder than on an animal-based diet mainly because plant sources are less dense, so more carbohydrate and fat ride along with each gram of protein.

• Judge any protein source by two questions: is it complete (does it carry all nine essential amino acids?) and how dense is it (how much protein per calorie, before carbs and fat crowd the plate)? Completeness matters over the course of a day; density needs to be considered in every meal.

• Those two answers sort every source into three roles: Anchors carry the meal (complete and dense), Bridges support it (dal, chana: real protein but carb-heavy), and Toppers garnish it (seeds, nuts: incomplete and calorie-dense).

Yes, vegetarians can hit the target

The evidence on whether vegetarian diets support muscle and protein needs has gotten cleaner in the last few years.

A 2023 RCT in young adults compared a high-protein vegan diet against a high-protein omnivorous diet over 10 weeks of progressive resistance training (Journal of Nutrition, 2023; n=22 in the training phase). Both groups gained lean mass at the same rate (+3.1 kg vegan, +2.6 kg omnivore), with comparable strength gains and identical daily muscle protein synthesis. A 2024 crossover trial in active older adults (average age 72) found the same: integrated daily muscle protein synthesis was identical on a 10-day vegan diet versus a matched omnivorous one, and the vegan diet also lowered LDL cholesterol (Journal of Nutrition, 2024; n=34). A South Asian-specific study found SA men gained the same muscle and lower-body strength from 12 weeks of strength training as White European men (Scientific Reports, 2022; n=34). The bodies respond. The plate has to deliver.

How much protein each meal needs to deliver depends on age. Older muscle is anabolically resistant: the same dose produces a smaller response, so the target per eating occasion rises with age, from about 0.24 g/kg in your 20s toward about 0.40 g/kg in older age (Moore et al., J Gerontol A Biol Sci Med Sci, 2015). In practical terms, that is 15 to 30 grams per eating occasion. This issue anchors on the top of that range, 30 grams, because it is calibrated for middle-aged readers, with room to scale to 35-40 g for those over 60 or training heavily. Landing protein in this window at each meal also blunts the post-meal glucose rise, a second reason to spread it across every meal rather than backloading it at dinner.

What that 30 grams looks like depends entirely on the source. A dal plate and a bowl of Greek yogurt can both deliver it. They are not the same plate.

How to judge any protein source

There are two questions to ask when evaluating any protein source:

1. Is it complete?

2. How dense is it?

Completeness

Your body needs nine essential amino acids it cannot make on its own. A “complete” protein source delivers all nine in proportions adequate for human needs. Animal proteins (dairy, eggs, fish, meat) are almost always complete. Most plant proteins are not. Legumes (dal, chana, rajma) are usually low in methionine and cysteine. Grains (rice, wheat) are usually low in lysine. Soy is the exception: it carries all nine in adequate proportions on its own, which is why tofu and soy chunks behave more like dairy than like legumes.

Here is where an old rule needs retiring. Many of us were taught that you must combine beans and rice in the same meal to make a complete protein. That idea, popularized in the 1970s, was walked back decades ago, including by the writer who first spread it. Your body keeps a circulating pool of amino acids and draws on it across the day, so completeness is a property of your day, not your plate. As our guest expert Ashley Koff, RD, puts it: the amino acids stay available across a 24-hour window, so what matters is that the day adds up to a full set, not that any single meal does. Eat dal at lunch and a cup of dahi at breakfast and you have already covered the gaps that dal-and-rice was supposed to close at one sitting. For a vegetarian eating any reasonable variety, completeness mostly takes care of itself.

One thing genuinely is a per-meal question, and it is leucine, not completeness. Leucine is the amino acid that triggers muscle protein synthesis directly, with the response strongest at roughly 2.5 to 3 grams in a single sitting, near the top of the acute leucine dose the ISSN position stand (2017) recommends. Unlike total amino-acid coverage, which your body assembles from its pool across the day, the muscle-building trigger responds to the dose in a single sitting. A 30-gram serving from dairy, eggs, or whey clears it easily; the same amount from soy or legumes sits closer to 2.3 to 2.5 grams, just at the line. This is the real reason to put a dense, complete anchor on the plate, and it is a dose question, not a pairing one.

Incomplete is not a dead end, and there are ways to handle it.

• Pair across the day. A legume and a grain cover each other’s gaps: dal is low in methionine and carries lysine, rice is the mirror image, and over a varied day the pairing happens without anyone planning it. Dal-chawal was solving this centuries before anyone studied protein completeness and scored different sources based on that (more below). The catch is doing it in one sitting: pairing fixes completeness, not the carbohydrate. Dal and rice is two carbohydrate sources stacked together, so the carb load climbs even as the amino acid profile fills out. That tension is the whole reason the South Asian plate runs the way it does.

• Top it with a complete, lean anchor. A scoop of whey or a pea-and-rice blend, a spoon of dahi, a few cubes of paneer alongside the dal clears the leucine threshold and adds almost no carbohydrate. This is the move that raises a meal’s protein quality without piling on the carbs.

• Supplement the limiting piece. For someone training hard and watching calories, free leucine or an essential-amino-acid powder hits the per-meal muscle trigger with no extra food. A precision tool, not an everyday requirement.

The FAO (the United Nations Food and Agriculture Organization) formalizes the completeness question with a score called DIAAS (digestible indispensable amino acid score), which captures both whether a source has all nine essential amino acids in adequate proportions and how much of it you can actually digest. Dairy, soy products, and eggs score “good” or “excellent”; lentils, chickpeas, kidney beans, wheat, white rice, and most seeds and nuts fall short on their own. In the table below, “complete” and “incomplete” stand in for those scores.

Density

Completeness is about amino-acid coverage. Density tells you how efficiently a source delivers: how much protein you get per calorie. This is the number that decides whether you can hit a protein target without a calorie flood, and it is the single most useful thing to know about a protein source. Greek yogurt hands you 30 grams of protein in 180 calories. Almonds, on the other hand, have 810 calories for the same 30 grams of protein. Same amount of protein, wildly different efficiency.

Density and completeness are independent, and that independence is what most vegetarian-diet advice misses. Full-fat paneer is complete but low-density: you pay for its protein in fat calories. Dal is incomplete and only moderately dense. A high quality score does not make a food an efficient protein, and a beloved staple is not automatically a lean one.

One more layer matters for a cardiometabolic reader: not just how many calories ride along, but what kind. Dal’s come as carbohydrate, which drives glucose and triglycerides. Full-fat paneer’s come as saturated fat, which drives ApoB and insulin resistance. The scorecard lists both so you can see which bill you are paying.

Three roles of a protein source

Put the two answers together and every protein source falls into one of three roles:

• Anchors are complete and dense. They carry the protein in a meal: dahi or Greek yogurt, whey, eggs, soy chunks, tofu, low-fat paneer, a pea-and-rice blend. A few anchors are complete but low-density, like full-fat paneer: real protein, but you pay in calories and saturated fat.

• Bridges are real protein but carb-heavy, eaten as a dish, and incomplete on their own: dal, chana, rajma. They support an anchor; they cannot replace one.

• Toppers are incomplete and low-density, a lot of calories for the little protein they add: seeds and nuts. They garnish a plate; they never anchor it.

The scorecard

Here is every common vegetarian protein, scored on both questions and sorted into its role. For each, this is what it takes to reach 30 grams of protein, and what comes along for the ride.

Anchors (complete; they carry the meal)

Bridges (incomplete alone; real protein, ride alongside an anchor)

Toppers (incomplete; sprinkle, don’t rely)

A few things jump out.

The lean anchors do the most work for the least. Greek yogurt, whey, egg whites, and a well-designed plant blend each deliver 30 grams of protein for under 250 calories and under 15 grams of carbohydrate. If you are trying to hit a high-protein target without doubling a meal’s calorie or carb load, these are the tools. Soy chunks (often sold as Nutrela in Indian grocery aisles) belong right with them and are the most underused anchor in South Asian vegetarian cooking. Their one tradeoff is carbohydrate, and it is worth being precise about: because textured soy is made from defatted soy flour, it keeps more of the bean’s soluble carbs than tofu does, so the same 30 grams of protein arrives with roughly four times the carbohydrate of tofu (about 19 grams versus 5). Still modest, and far below dal.

Among the powders, the choice is mostly about source, not efficiency, since all of them are lean. Whey and soy isolate are complete proteins on their own. A pea-and-rice blend is complete by design, the rice supplying the methionine that pea is short on. Pea protein alone is the one to read the label on, because without that rice it stays low in methionine (the lone ⚠️ among the powders in the table).

Full-fat paneer is the cautionary anchor: complete protein, but 42 grams of fat and 27 grams of saturated fat for that 30 grams, more saturated fat than most days’ recommended limit in a single serving. Fine occasionally, expensive daily, and a real problem if your ApoB is climbing.

Then the bridges. Dal’s density is only moderate, and the real problem is what fills it out: you eat 67 grams of carbohydrate to get 30 grams of protein, roughly half a day’s budget before any rice or roti. By the macros, dal is a carbohydrate source that contains protein. Chana is worse (90 grams of carbohydrate for the same 30 grams of protein).

Seeds and nuts sit at the bottom as toppers. Hemp hearts are the closest to complete, but reaching 30 grams from any seed or nut costs 500 to 800 calories and 45 to 70 grams of fat. A tablespoon of hemp on dahi or a handful of pumpkin seeds adds protein and useful minerals; no one is building a meal’s protein around almonds.

A few caveats. The RCTs showing equivalent muscle adaptations are mostly short (10-12 weeks) and mostly in younger adults; long-term data in older vegetarians is thinner. The 30 g per-meal target is a mid-range calibration; heavier anabolic resistance may push it to 35-40 g. The DIAAS values for paneer are functional estimates, not formal assays, and seeds and nuts are marked incomplete rather than scored because published values are for protein isolates, not the whole food. The direction holds; the precision at the edges does not.

The anchor-first plate

A 100-gram protein day on a vegetarian plate is mostly a question of architecture, not new recipes. The rule is simple: an anchor at every meal, with bridges and toppers in support.

Breakfast. A cup of high-protein dahi (look for 15-17 g per cup and check the protein-to-carb ratio on the label) or a small protein smoothie is the easiest single move. No cooking change, no household disruption.

Lunch. A dal-and-vegetable plate with a real anchor on top: a side of tofu sabzi, a few cubes of paneer, two eggs if you eat them, or a generous bowl of plain Greek yogurt alongside. The dal stays as the bridge; the anchor sits next to it.

Dinner. A protein anchor (paneer, tofu, soy chunks, or eggs) plus dal plus vegetables plus a measured carbohydrate. The carbohydrate becomes the smaller player on the plate, not the largest.

That sequence assembles roughly 90 to 110 grams of protein across three meals, which lands in the 1.2-1.6 g/kg range for most adult readers.

Myth: “Dal is high in protein”

The Myth: Indian households have called dal “protein” for generations. Reddit threads, family WhatsApps, and primary care doctors repeat it. If you eat dal regularly, the assumption is that the protein box is checked.

The Evidence: The scorecard above settles it. By the macros, dal is a carbohydrate that carries some protein, not a protein source, once you see the carb load it takes to reach 30 grams. Its quality is modest too: on DIAAS, dairy, soy products, and eggs clear the FAO “good” threshold; lentils, chickpeas, and kidney beans do not.

Dal is a bridge, not an anchor. It is real protein and a useful support food in a well-designed plate, but the anchor has to come from somewhere else.

Questions you’re probably asking

“I always eat dal with rice because I was told the combination makes a complete protein. But my triglycerides keep climbing. Am I doing something wrong?”

First, the good news: you can stop worrying about the completeness part. The rule that you must pair beans and rice to build a complete protein has been retired. Your body keeps a pool of amino acids and draws on it across the day, so a varied vegetarian diet covers the gaps on its own. You were never at risk of an incomplete plate. What the pairing does do is stack two carbohydrate sources together, and that is where the triglycerides come in. A dal-rice-roti plate is carbohydrate-dominant, and in insulin-resistant South Asians a carbohydrate-heavy meal pushes the liver to make more VLDL, the particle that ferries triglycerides into your bloodstream. That is the climb showing up on your panel, and a normal protein total does nothing to flag it. The fix is not to give up the flavors. It is to let the protein come from a lean anchor at each meal and let the carbohydrate be the smaller player. The carbohydrate load, and the triglycerides, come down with it.

“I have been vegetarian my whole life and just started lifting. I eat plenty of dal, but I am not gaining strength the way my gym friends are. Is it the diet?”

Usually a per-meal leucine gap, not a total-protein problem. A dal-anchored meal often falls short of the leucine dose that flips the muscle-building switch. The fix is not more dal. It is a leucine-rich anchor at each meal: high-protein dahi, soy chunks, paneer, eggs, or a scoop of whey or plant blend. The other lever, creatine, gets its own note just below. Same training, different plate.

Vegetarians have options beyond dal. One just has to be deliberate about how to construct the plate, and use data to inform your approach.

Ask Your Doctor About: Creatine

If you are a lifelong vegetarian who lifts, this is the one supplement with a clear, specific payoff. Creatine comes almost entirely from animal foods, so vegetarians carry lower muscle creatine stores to begin with, and in a controlled trial they gained more lean mass from creatine plus resistance training than omnivores on the same protocol (Burke et al., Medicine & Science in Sports & Exercise, 2003). Five grams a day of creatine monohydrate is the standard maintenance dose and a sensible floor; larger bodies and heavier training loads often do better on more. Cheap, well-studied, and the clearest vegetarian win in this whole issue.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

When people say “strength training,” they usually lump together three different goals: getting stronger, getting bigger, and building endurance. They sound like one thing. In reality, they’re trained differently, and for your health they aren’t equally worth your time.

The difference matters more as you age. You have to be a lot more focused, and you have to make sure you recover well. In this post, we get into the details of the goals, offer a perspective on which ones actually move the needle, and show how to build a week around them.

TL;DR

• Resistance training builds three things: strength (force), hypertrophy (size), and endurance (fatigue resistance). They’re trained with different loads, but effort matters more than the exact rep number.

• For metabolic health, hypertrophy is the priority. Muscle is your largest glucose sink, and building it barely depends on the “right” rep range, just hard sets.

• After 40, add strength and power (for function, balance, and avoiding falls) and some heavier, higher-impact loading (for bone). Endurance is the least essential of the three.

• The week is simpler than it sounds: two or three short sessions, most sets 1 to 3 reps short of failure, and a heavier or faster day once you’re past 40.

Three goals, trained three ways

Resistance training builds three distinct qualities. Strength is how much force you can produce. Hypertrophy is the muscle physically getting bigger. Local muscular endurance is how long a muscle can keep working before it fatigues.

One nuance sits inside strength. Power is how fast you can produce force, the speed layered on top of raw strength: a powerlifter grinding out a heavy deadlift is expressing strength, while a sprinter exploding out of the blocks is expressing power. Most people don’t train the two separately, which is why the classic gym chart lists three, not four. That chart assigns each a rep range: low reps for strength and power, moderate for size, high for endurance (Schoenfeld et al., Sports, 2021; a review re-examining that “repetition continuum”). It’s a useful starting point, and looser than it looks.

Here’s what the strongest evidence shows. For raw strength, load is the lever: heavier weights, above about 80% of your one-rep max, produce the biggest gains. For size, the load barely matters. Light, moderate, and heavy all grow muscle to a similar degree, as long as the sets are hard (Currier et al., British Journal of Sports Medicine, 2023; 178 strength studies with 5,097 people, and 119 muscle-size studies with 3,364). An earlier review found the same: taken close to failure, light and heavy loads built muscle about equally (Schoenfeld et al., Journal of Strength and Conditioning Research, 2017; 21 studies).

What ties them together is effort, not the rep count. The useful gauge is reps in reserve, which means how many more you could have done. You don’t need to grind to failure. Stopping 1 to 3 reps short captures nearly all the muscle growth (Refalo et al., Sports Medicine, 2023; systematic review and meta-analysis). Going closer to failure does nothing extra for strength (Robinson et al., Sports Medicine, 2024; meta-analysis, 55 hypertrophy and 67 strength studies), but it does raise the injury risk.

What your metabolism needs, and what changes with age

Building muscle is the metabolic priority, because muscle is your largest glucose sink. Most of the sugar in your blood gets cleared into it, so more muscle is better for metabolic health. We walked through that mechanism in our training week issue. The form of training that matters most for your metabolism is hypertrophy, and because size responds to hard sets across a wide rep range, you don’t need a perfect rep number to get it. This matters most for people who look fine on the scale but carry the thin-fat pattern, common in South Asians, where a normal weight can hide too little muscle.

After about 40, two more things start to count: power and bone density. Power, the ability to produce force quickly, fades faster than raw strength and tracks more closely with everyday function, whether you can climb stairs, catch a stumble, or get out of a chair. Training that moved weights quickly was associated with a modest improvement in physical function in older adults compared with slower strength work (Balachandran et al., JAMA Network Open, 2022; systematic review and meta-analysis).

Bone responds to heavier, higher-impact loading. In postmenopausal women with low bone mass, twice-weekly high-intensity resistance and impact training raised lumbar-spine bone density by about 4% while a gentler program did not (Watson et al., Journal of Bone and Mineral Research, 2018; randomized controlled trial, n=101). That kind of training has to be coached and built up gradually. The trial stayed safe precisely because it was supervised and progressive.

That leaves endurance. High-rep, light-load work builds a muscle’s resistance to fatigue, which is useful if you enjoy it, but it’s the least essential of the three for staying metabolically healthy and independent.

How to build a week around it

Putting it together is simple. The Best Training Week Is the One You Repeat issue laid out the full picture: two or three strength sessions, some easy cardio, a daily walk. Here’s how to spend the strength part so it covers what matters.

On each strength day, work the major muscle groups, push, pull, hinge, and squat, in a range that’s genuinely challenging, roughly 6 to 12 reps, with most sets stopped 1 to 3 reps short of failure. That one choice covers hypertrophy and a good share of strength at the same time. Let the weights climb over the months. That progression is the signal your body actually responds to.

If you’re past 40, add two small things. One heavier day, where you drop to about 5 or 6 hard reps on the main lifts, builds strength and loads the bone. And a few fast, controlled reps, moving the weight with intent, train power. Neither needs its own session; they fold into the days you already have.

You can skip dedicated endurance work. High-rep burnout sets aren’t where your health payoff is, and the challenging sets you’re already doing build plenty of staying power.

A simple check that a session counted: the last couple of reps of each set were hard, your form held, and you’re lifting a little more than you were a few months ago.

The three forms aren’t a test you have to ace. For your health, the order is important: build muscle first, your metabolic engine; keep strength and power as you age, your independence; treat endurance as a bonus, not a goal. Get that order right, push your sets close enough to count, and the exact rep range you used to get there stops being a question worth asking.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

This week’s issue is co-authored with Dr. Tiffany Mullen, DO, a functional medicine physician and founder of Vytal Health.

In Auckland, New Zealand, 84% of South Asian women tested below the vitamin D sufficiency line, with a median level of just 11 ng/mL. That is deep deficiency, in a country famous for its sun.

My family in India, one of the sunniest places on earth, runs low too. I live in the Bay Area in California (plenty of sun), and I run low without high dose supplementation. If you are South Asian, this probably sounds familiar - a parent put on vitamin D after a fracture scare, or a doctor who mentioned in passing that people like us are usually deficient. The pattern is common: the people with the most sun overhead were the most deficient inside.

That is the paradox I want to walk through. Because once you understand why it happens, the standard advice your doctor gives you, take 600 to 800 IU and you are covered, stops making sense for people with dark skin.

TL;DR

• South Asians are among the most vitamin D deficient people on earth, even in sunny climates: in one Auckland study, 84% of South Asian women fell below the sufficiency line (Public Health Nutrition, 2010; cross-sectional, n=228).

• The standard recommendation of 600 to 800 IU a day was built to protect bone in mostly light-skinned populations, and it routinely underdoses South Asians.

• In South Asian women, insulin resistance only improved once vitamin D climbed past 32 ng/mL, well above the “adequate” line of 20 (British Journal of Nutrition, 2010; randomized trial, n=81).

• Vitamin D is a hormone with both a floor and a ceiling: correcting a real deficiency helps, but chasing very high levels backfired in trials (more falls, less bone), so aim for a range rather than a maximum. South Asians are far more likely to be starting below it.

The sun can’t fix this for us

Start with how strange the data is. In India, Pakistan, and Bangladesh, countries with abundant year-round sun, cohort after cohort puts roughly 70 to 80% of adults below 20 ng/mL, the line where most labs call you deficient (review of South Asian studies, Pakistan Journal of Medical Sciences, 2008). In Australia’s national survey, 67% of people born in Southern and Central Asia were deficient, and 60% stayed deficient even in summer (Australian Bureau of Statistics, 2011-12 National Health Measures Survey).

Five things stack up against us, and they compound.

• Skin. Melanin is natural sun protection, so darker skin makes far less vitamin D per minute of sun, several times less than fair skin. (How many minutes that actually costs you is the next section.)

• Body fat. Vitamin D is fat-soluble, so it gets pulled into fat tissue and out of circulation. South Asians carry more body fat at a given weight, so more of the vitamin gets stored instead of used, and people with higher body fat often need two to three times the dose to reach the same blood level (Endocrine Society guideline, JCEM, 2011).

• Sun avoidance. Indoor work, plus a cultural premium on staying fair, keeps a lot of skin covered, especially for women.

• Diet. Few foods carry much vitamin D, and UK South Asians average just 1 to 3 micrograms a day from food, a fraction of what you need (UK Biobank data, Proceedings of the Nutrition Society, 2020).

• Genetics. Common variants in the genes that make and carry vitamin D leave some people lower for the same sun and the same dose.

Put them together and casual sun plus a token supplement does not move the needle.

On sun alone, 25 minutes a day, and not in winter

The obvious fix is to just go outside, so it is worth knowing the actual number. For South Asian skin, it takes roughly 25 minutes of midday sun with your forearms and lower legs bare, most days, to hold a healthy level. Fair skin gets there in about 9 minutes (Webb et al., Nutrients, 2018). The melanin that protects you is the same thing that roughly triples the clock.

And those minutes only count under real conditions: midday, when the UV index is above 3, with no sunscreen (an SPF of 8 blocks most of the rays you need) and arms and legs actually bare, not just hands and face. Above about 35 degrees of latitude, which covers most of the United States, Europe, and northern India, the winter sun sits too low to make much vitamin D at all, for months at a stretch (Holick, NEJM, 2007). The one piece of good news: you cannot overdose this way, because your skin stops once it has made enough.

So yes, the sun can do it, at midday, in the sunny months, with your limbs bare, for 25 minutes most days. For most of us, most of the year, that is not a plan you can count on. Which is why you supplement to a target instead. The question is what that target should be.

20 is a bone number, not a metabolism number

For anyone watching their metabolic health, the standard target is set too low, and here is why.

The “adequate” line most labs use, 20 ng/mL, comes from the Institute of Medicine. It was set to prevent rickets and protect bone, modeled mostly on light-skinned populations, assuming some sun and decent calcium intake. It was never designed around insulin, diabetes, or heart disease. It answers one question: are your bones covered?

Your metabolism appears to want more. In a randomized trial, 81 insulin-resistant South Asian women in New Zealand took either 4000 IU of vitamin D a day or a placebo for six months. The supplemented group improved their insulin sensitivity and dropped their fasting insulin, but the benefit only showed up once blood levels passed 32 ng/mL, with the best range landing between 32 and 48 (British Journal of Nutrition, 2010; randomized trial, n=81). At 22 ng/mL your lab says “sufficient” and your doctor moves on. Your bones may be fine there. The metabolic benefit in these women did not appear until they were ten points higher.

But the big trials look mixed

The trials on this are genuinely mixed, and you should know it. When researchers pooled 18 randomized trials, vitamin D supplementation showed no overall effect on insulin sensitivity (Diabetes Care, 2020; meta-analysis, n=1,220), and the largest diabetes-prevention trial gave 4000 IU a day to people with prediabetes and saw no significant drop in who progressed to diabetes (NEJM, 2019; randomized trial, n=2,423).

So which is it? The two findings reconcile cleanly. Vitamin D behaves like a floor, not a lever. Climb out of deficiency and you may gain something; keep adding once you are already sufficient and nothing much happens. That diabetes-prevention trial never screened for deficiency, so most people started near 28 ng/mL with no floor to climb off, while the South Asian trial enrolled women who were deficient and insulin resistant and pushed them past the threshold. The benefit lives in the climb out, and South Asians, already wired toward insulin resistance at lower body weights, are far more likely to be starting down there. It is the same thread I pulled on in the fasting insulin issue: the standard target is aiming at the wrong number for the wrong population.

Vitamin D is a hormone, and hormones have a ceiling

So your metabolism wants more than the standard target. There is an upper bound too, though, and finding it changed how I think about my own dose. Vitamin D is not really a vitamin. Your body converts it into calcitriol, a steroid hormone from the same family as cortisol and the sex hormones, and it behaves the way hormones do, switching genes on and off and controlling how much calcium moves through your blood. Calling it a vitamin makes it sound like vitamin C, where the excess just leaves in your urine. It does not. It is fat-soluble, it accumulates, and it acts through a receptor.

So the dose has a ceiling, and you can overshoot it. Push your level too high and you get too much calcium in the blood and urine, which can mean kidney stones and, the part that matters for anyone watching their heart, calcium settling into soft tissue like artery walls.

The trials are blunt about this. When more than 2,200 older women were given a single 500,000 IU dose of vitamin D once a year, they had 26% more fractures and more falls than the placebo group, with the spike worst in the first three months (Sanders et al., JAMA, 2010; randomized trial, n=2,256). And when healthy adults took 400, 4,000, or 10,000 IU a day for three years, the higher doses did not build bone, they lost slightly more of it, in a clean dose-response (Burt et al., JAMA, 2019; randomized trial, n=311). In both trials, the higher the dose, the worse the result.

So where is the real ceiling? Higher than the 20 most labs fixate on. The range most functional-medicine clinicians aim for is 40 to 60 ng/mL, where the benefit is best supported and the risk stays low. Above 60, the added benefit is unproven for general health, and the highest-dose arm of that bone trial, the group on 10,000 IU a day whose levels climbed into the 70s, lost the most bone. Frank toxicity, the hypercalcemia kind, is rarer still and shows up far higher, north of 100. None of this is a reason for fear. It is a reason to aim for a range: correct a deficiency, settle into the 40s or 50s, and stop chasing a bigger number.

From Amandeep: I take 5000 IU a day, which is above the 4000 IU daily ceiling most guidelines set for routine use, and I do it on purpose. My genetic panel flagged a variant in my vitamin D receptor, the docking station the active hormone binds to (the gene is called VDR). My conversion genes are normal, so for me the bottleneck is not making or storing vitamin D, it is the receptor reading it. That is part of why the same dose lands two people at completely different blood levels: your genes, your body fat, and your sun exposure all move the result. The science tying a specific receptor variant to a specific dose is still emerging, so I treat it as one input, not a verdict. It nudged my personal target higher, to 45 to 65 ng/mL, and my last test landed at 55 (April 2026), inside the healthy range but on a bigger dose than most people need. Because I run that high a dose, I track my blood calcium (normal so far) and pair the D with vitamin K2 (the MK-7 form), which helps steer calcium toward bone instead of arteries.

Myth: “Brown skin adapted to the sun, so I get plenty of vitamin D”

The Myth: People with South Asian skin evolved under strong sun, so we are naturally good at making vitamin D and do not need to worry about it, especially if we live somewhere warm.

The Evidence: The exact opposite is true, and the mechanism is the giveaway. Melanin protects skin by blocking some of the UVB that triggers vitamin D production, so darker skin makes less per minute of sun, not more. That is why 84% of South Asian women in sunny Auckland tested deficient (Public Health Nutrition, 2010; n=228), and why 60% of Southern and Central Asian-born Australians stayed deficient through summer (ABS, 2011-12). Add indoor lives and clothing that covers arms and legs, and the one advantage, abundant sun, never reaches the skin.

The pigment that protects us from the sun is the same pigment that keeps us from making vitamin D from it. Warm weather does not fix this.

From Amandeep: One Thing I Changed

I stopped treating my vitamin D dose as fixed and started treating it as a dial. I moved my testing to a regular cadence and adjusted the dose against the result, the same way I would tune any other marker. The point was never to hit a big number. It was to land in a range on purpose instead of hoping a standard dose was enough.

Vitamin D is a deficiency worth correcting, especially in a population that is deficient by default and metabolically vulnerable on top of it. That is a modest, well-supported claim, and it is enough.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

Spoiler alert: the part that has the biggest impact is consistency, and how the week is spaced. The metabolic benefit of a single workout, the improvement in how your body handles glucose, fades within about 72 hours. Go four or more days without anything and you drift back toward where you started. A consistent routine you follow week after week, without long gaps, beats the cleverest program you follow for a month and quit.

This post is the whole picture in one place. What strength training does, what cardio does, why walking is the floor and not the workout, and how to assemble it into a week you can sustain. We wrote it with Trei Tackett, MSN, FNP, an executive health coach and founder of Pillars Health, a precision medicine company, who programs this for a living and has coached hundreds of people over the years.

Two things are non-negotiable, and you need both: strength and cardio. They do different jobs, and done together they beat either one alone. Combined training is associated with about 40% lower all-cause mortality than strength or cardio by itself (British Journal of Sports Medicine, 2022), and a year-long trial of 406 adults found that splitting time evenly between the two matched a full cardio program for cardiovascular risk while adding the muscle that cardio alone misses (European Heart Journal, 2024; randomized controlled trial, n=406). We’ll take them one at a time, then put the week together.

TL;DR

• Strength training is non-negotiable. The mortality benefit maxes out at just 30 to 60 minutes a week, and skeletal muscle handles about 80% of the glucose your body clears after a meal

• Cardio is non-negotiable too. Low cardiorespiratory fitness is comparable to smoking as a risk factor for early death, and fitness is the single biggest lever on South Asian insulin resistance

• Walking is the floor your training stands on, not the workout itself. 7,000 steps a day and breaking up long sitting both matter on their own, before any workout

• Never let three days pass without a session. Consistency and spacing beat any specific program

Your muscles run your metabolism

Of the two, strength is the one people skip, and it carries the most leverage. Strength training is associated with 10 to 17% lower risk of all-cause mortality, cardiovascular disease, cancer, and diabetes, and the benefit maxes out at just 30 to 60 minutes a week (British Journal of Sports Medicine, 2022; meta-analysis, 16 studies, n=263,058). That is less time than most people spend choosing a program.

And the case only gets stronger with age, because muscle doesn’t hold steady on its own. After about 30 we start losing it, roughly 3 to 8% per decade, with the decline accelerating after 60 (Frontiers in Physiology, 2012; quantitative review of skeletal muscle aging). Left unchecked, that’s muscle quietly traded for fat, less strength, and a harder time staying independent in your 70s and 80s. Resistance training is the one reliable way to build muscle and hold onto it. Cardio won’t do it, and neither will diet alone.

Muscle is also far more than structural. We tend to think of it as the stuff that moves the skeleton around, but it’s the largest metabolic organ you have. Skeletal muscle is responsible for roughly 80% of the glucose your body clears from the blood after a meal under the influence of insulin (Journal of Clinical Investigation, 1985; DeFronzo et al., euglycemic clamp study). Not the liver, not fat tissue. Your muscles are the primary place sugar goes, and more muscle means a larger, more responsive sink for it. In a large national sample, every 10% increase in relative muscle mass was associated with an 11% lower measure of insulin resistance and a 12% lower rate of pre- or overt diabetes, even after accounting for overall obesity (Journal of Clinical Endocrinology & Metabolism, 2011; Srikanthan and Karlamangla, NHANES III, n=13,644).

This is where South Asians start at a disadvantage. At the same body fat, South Asian men carry about 3.5 kg less lean mass than European men (Journal of Clinical Endocrinology & Metabolism, 2009; Lear et al., n=828). A UK Biobank imaging study found the gap holds even when you match people for sex, age, height, and weight: South Asians had less muscle and more fat, and more than twice the rate of poor muscle composition, 19.9% versus 7.9% (eBioMedicine, 2025; UK Biobank imaging cohort). The “thin-fat” pattern, low muscle paired with high visceral fat, is common in South Asians, and one South Indian cohort put the odds of type 2 diabetes at close to 2.7 times higher even at a normal weight (PLOS One, 2020; Kerala Diabetes Prevention Program baseline, n=1,147). The scale can look fine while the body composition underneath it isn’t.

The good news is that muscle is built the same way in everyone, and it doesn’t take much equipment or time.

From Amandeep: Between July 2024 and January 2025, my DEXA scans showed something I had to read twice. My body fat dropped from 39% to 30.7%, but my lean mass went up, from 141 to 150 pounds. I gained 9 pounds of muscle while losing fat. Visceral fat fell from 5.5 to 3.89 pounds, a 29% drop. The combination that drove it: a GLP-1 medication, strength training three times a week, and high protein. The lifting was what turned weight loss into the right kind of weight loss.

That last point matters more than ever now that so many people are on GLP-1 drugs like tirzepatide (Mounjaro, Zepbound) and semaglutide. These drugs work, but what you lose depends on what you do alongside them. In the SURMOUNT-1 body-composition substudy, tirzepatide produced 21% weight loss, of which roughly a quarter was lean mass, not fat (Diabetes, Obesity and Metabolism, 2025; n=160 with DXA). For South Asians, who start with less muscle, losing a quarter of it is a worse trade than it sounds.

It doesn’t have to go that way. In a randomized trial, adding a structured exercise program to GLP-1 therapy roughly doubled fat loss while preserving lean mass, compared with the medication alone (New England Journal of Medicine, 2021; Lundgren et al., randomized controlled trial, n=195). Training is what shifts more of the weight you lose toward fat and away from muscle. The minimum worth holding to on these medications: lift at least twice a week, and aim for 1.2 to 1.6 grams of protein per kilogram of body weight per day.

How much strength training is enough, and what kind? Less than you’d guess, and simpler than the internet makes it. The American College of Sports Medicine published its first resistance-training update in 17 years, pulling together 137 reviews and more than 30,000 participants (Medicine & Science in Sports & Exercise, 2026). The headline: the biggest jump comes from going from no resistance training to any. From there, a few things reliably drive strength: lifting heavier loads (around 80% of your one-rep max, heavy enough that the last couple of reps are genuinely hard), through a full range of motion, two to three sets, at least twice a week, done early in the session when you’re fresh. What surprisingly didn’t move the needle: training every set to failure, the type of equipment, how complex the exercises were, or any elaborate periodization scheme. Bands, bodyweight, and home routines all count. You do not need a barbell. You just need to keep showing up.

Cardio is where fitness is won

Where strength protects your muscle and metabolism, cardio protects your heart and your staying power. In a study of 122,007 patients followed for 8.4 years, low cardiorespiratory fitness was comparable to smoking and diabetes as a risk factor for early death (JAMA Network Open, 2018; retrospective cohort). Moving from the low-fitness group to just below average cut mortality risk by about half. Fitness here is VO2 max, how efficiently your body uses oxygen when you push it. You don’t need to measure it to benefit. You need to do cardio regularly.

And fitness isn’t only about living longer. It’s what lets you climb stairs, carry your bags, and keep up on a trip without getting winded at 60 and 70. The more aerobic capacity you build now, the more of it you keep as the years take their cut.

For South Asians, one number reframes the whole conversation. In a comparison of 100 South Asian and 100 European men matched for BMI and age, lower cardiorespiratory fitness explained 68% of the excess insulin resistance in the South Asian group (Diabetologia, 2013; Ghouri et al., n=186). Fitness here isn’t separate from muscle. It reflects how well your muscles and their mitochondria burn fuel, and you raise it by moving more. But the same researchers found something we can’t easily train our way around: South Asian men were less fit than Europeans at every level of physical activity, not because they moved less. The gap appears built into how the muscle handles fuel, which means we may need to train more than the standard guidelines assume to land in the same place. For us, that makes fitness the single highest-return thing to train.

Cardio comes in two gears, and you want both. The easy gear is what’s usually called Zone 2: a conversational pace you could keep up for an hour, breathing a little harder but still able to talk in full sentences. It builds your aerobic base, trains your muscles to burn fat, and costs little to recover from, which is why it should be the bulk of your weekly cardio.

The hard gear is intervals: short bursts of near-maximal effort with easy recovery between them. They raise fitness faster per minute than anything else you can do. In a 12-week trial, 10 minutes of sprint intervals done three times a week produced the same gains in fitness and insulin sensitivity as 45 minutes of Zone 2 cardio (PLOS ONE, 2016; randomized controlled trial). That efficiency is real, but intervals are the seasoning, not the meal. A little goes a long way, and more is not better.

You don’t need a special name or machine. A few interval protocols have the deepest evidence behind them, and the honest truth is that the one you’ll repeat beats the one that’s theoretically optimal.

• The Norwegian 4x4, the most studied of them: four rounds of four minutes hard with three minutes easy between, about 30 minutes total.

• Sprint intervals, the highest return per minute: three 20-second all-out efforts with two minutes of easy recovery, roughly 10 minutes total.

• Tabata, tighter and genuinely hard: 20 seconds hard, 10 seconds easy, eight rounds, four minutes.

Run hills, ride a bike, row, swim, jump rope. The modality matters less than picking something you’ll come back to next week.

If you think high intensity is risky for regular people, the data says otherwise. In supervised programs for patients who already had cardiovascular disease, there was one major cardiac event for every 17,083 sessions (Journal of the American Heart Association, 2018; systematic review). If it’s safe enough for people with heart disease under supervision, most of us can ease into it.

The mistake people make is doing too much of it. Intervals are the most taxing thing per minute that you can do, and more is not better. A small 2025 trial found that two HIIT sessions a week produced the same cardiovascular benefit as three (Physiological Reports, 2025; randomized controlled trial, n=26). Past two or three sessions, recovery cost climbs, cortisol stays elevated, sleep can suffer, and chronically high cortisol can add visceral fat, the opposite of the point. One interval session a week is a great start. Two is the sweet spot for most people.

Intervals also hit two things South Asians tend to carry more of. Running-based interval training reduces visceral fat, the kind packed around the organs, and a review of 19 studies found it cut liver fat by about 2.85%, in less time than Zone 2 cardio takes to do the same (Journal of Clinical Endocrinology & Metabolism, 2022; meta-analysis, n=745). Fatty liver affects roughly a third of South Asians, so that is not a small side benefit.

There’s a reason intervals stick as a habit once you start. Hard efforts trigger a surge in BDNF, the brain protein most tied to focus and learning, along with norepinephrine and endorphins. The heart benefits are the reason to do it. The clear-headed, slightly wired feeling afterward is the reason you’ll want to.

Walking is the floor, not the workout

Walking is baseline movement, the floor that your training stands on top of. It’s how a body is supposed to move through a day, and it sits underneath everything else you do. It’s essential, but it’s movement, not training: it doesn’t load your muscles or tax your heart the way strength and cardio do.

The largest analysis to date found that 7,000 steps a day was associated with 47% lower all-cause mortality, along with lower rates of heart disease, diabetes, dementia, and depression (Lancet Public Health, 2025; meta-analysis, 57 studies). Even 4,000 steps clearly beat 2,000. This is the base your training week stands on, not a substitute for it.

And sitting all day, then doing an hour in the gym, doesn’t undo the sitting. Two-minute walking breaks every 20 minutes during long stretches of sitting cut the glucose response to a meal by 24%, whether or not you also exercised that day (Diabetes Care, 2012; n=19). Prolonged sitting is its own risk factor, separate from how much you train.

One specific habit is worth building, and a lot of our families already have a version of it: a 10 to 15 minute walk after meals. The after-dinner stroll that so many South Asian households grew up with turns out to be better timed than it looks. A review found that walking after eating blunted glucose spikes more than walking before eating, which had almost no effect on the post-meal rise (Sports Medicine, 2023; systematic review). If you wear a continuous glucose monitor, you can watch it happen in real time. The timing is the trick: after the meal, not before.

What a week looks like

Put it together and the picture is simpler than the fitness industry makes it sound.

The minimum that gets you most of the benefit is about two and a half hours a week: two strength sessions of 30 minutes hitting the major muscle groups, one Zone 2 cardio session of 30 to 60 minutes, and 7,000 or more steps on most days. The jump from doing nothing to doing this is the single biggest improvement you can make. Everything beyond it has smaller returns.

When you’re ready to add, here’s the order. A third strength day. More Zone 2 cardio: build toward about 150 minutes a week, spread across two or three sessions, because one session a week barely holds your fitness in place while two or three is where it actually climbs. Then a single interval workout in place of one of those Zone 2 sessions. Ten or fifteen minutes of mobility work after your lifts. Keep intervals capped at one or two sessions a week, because they cost the most to recover from.

Here’s one way a fuller week can come together:

• Monday: strength, full body, 40 minutes

• Tuesday: Zone 2 cardio, 30 to 45 minutes

• Wednesday: strength, full body, 40 minutes, then 10 minutes of mobility work

• Thursday: Zone 2 cardio, 30 to 45 minutes

• Friday: strength, 40 minutes

• Saturday: one interval session, 10 to 20 minutes, plus the longest walk of the week

• Sunday: rest, some mobility, and a walk

That’s three strength days, three cardio sessions with only one of them intense, steps every day, and Sunday as a true rest day. The one tweak Trei makes for most clients is exactly that Thursday slot: a second Zone 2 session, which nudges your weekly aerobic volume toward the 150-minute mark without adding another hard day. Nothing in the week asks for more than 40 minutes at a stretch, and no two days lean on the same muscle groups back to back.

The one principle that runs through all of it: your right amount depends on everything else going on. Someone eating in a deficit and lifting heavy should lean toward Zone 2 over intervals, because the recovery budget is already spent. Stress, sleep, and life load all factor in. The best week is the one you can repeat when work gets busy, not the one that looks best on paper.

After 40, recovery earns its place

Two things change as you get older, and neither is what most people fear.

The first is recovery. Training status matters more than age: a well-trained 50-year-old recovers faster than an untrained 30-year-old. But the spacing does matter. Leave 48 to 72 hours between hard sessions for the same muscle groups, and treat that gap as part of the program, not a break from it.

The second is mobility and balance, and this is the part that’s easy to skip because the payoff is decades out. Falls are what take people out of independent life in their 70s and 80s. Hip mobility, shoulder range, single-leg balance: the work that prevents that starts now, not at 65. Ten or fifteen minutes after a strength session is enough to start.

What doesn’t change with age is the value of intensity, and that’s the reason not to drift into easy-only training. In a Mayo Clinic trial, 12 weeks of interval training raised mitochondrial capacity by 69% in adults aged 65 to 80, and reversed some of the age-related decline in how muscle builds protein (Cell Metabolism, 2017; randomized trial, n=72). The younger group in the same study gained less. Used in the right dose, hard effort is one of the few things that pushes back on cellular aging, which is exactly why it belongs in the week at 50 and 60, not just at 30.

All of this adds up to one thing that matters more every year: staying injury-free. An injury is the fastest way to break the consistency the whole program depends on, and a tweak you’d have shrugged off at 30 can cost you weeks at 55. The protection is built into the week itself: strength, mobility, and recovery each guard against injury, which is reason enough to keep all three. Beyond that, don’t overtrain, and listen to your body. A nagging ache that won’t settle is information and your body is telling you to back off, let it heal, but keep moving as long as there’s no pain. The goal was never the hardest week. It’s all about consistency and long term compounding.

Where to start this week

We asked Trei to put together a starting routine for someone beginning from scratch. It’s one full-body session, and you do the whole thing on each of your strength days. It needs nothing but your bodyweight or a set of resistance bands. Move through all four exercises in order, resting 60 to 90 seconds between sets, at a controlled tempo: about two seconds to lower, a brief pause, and one to two seconds to lift. The last two or three reps of each set should feel challenging, but never so hard that your form breaks down.

• Push (chest, shoulders, triceps): Push-up or dumbbell press, 3 sets of 10. Keep your body in a straight line from head to heels and your core braced. Lower until your elbows reach about 90 degrees, then press back up. If a floor push-up is too hard, elevate your hands on a counter or bench to reduce the load.

• Pull (upper back, biceps, rear shoulders): Resistance band row or dumbbell row, 3 sets of 10. Anchor the band at chest height or hinge forward with a dumbbell. Pull your elbows back and down, squeezing your shoulder blades together at the end of each rep. Don’t shrug your shoulders toward your ears.

• Hip hinge (hamstrings, glutes, lower back): Romanian deadlift or hip bridge, 3 sets of 10. For the deadlift, push your hips back with a soft bend in the knees and a flat back, lowering until you feel a gentle stretch in the hamstrings. For the bridge, drive through your heels and squeeze your glutes at the top without arching your lower back.

• Squat or lunge (quads, glutes, core): Goblet squat or split squat, 3 sets of 10. Hold a weight at your chest and sit back and down, heels planted and chest tall. For the split squat, finish all 10 reps on one leg before switching, lowering your back knee toward the floor under control.

Progression: Once you can finish all three sets of 10 with clean form and two or more reps left in the tank, make it harder by adding resistance, slowing the tempo, or adding a fourth set. Build gradually, not all at once.

Trei’s recommendation: Put those two days on the calendar like a standing meeting, same days, same time. Don’t rely on motivation or willpower. It’s much more reliable to build systems. The clients who succeed long-term aren’t the most disciplined. They make training a default instead of a daily decision. James Clear’s Atomic Habits gives the test for a habit that holds: make it obvious, attractive, easy, and satisfying. Obvious = scheduled. Attractive = keep it light, don’t crush yourself. Easy = a simple plan, movements you’re familiar with. Satisfying = enjoy the mood boost, or a small reward from the smoothie bar when you’re done.

If you take one thing from all of this, it’s that consistency beats optimization, every time. Build the week you can actually repeat, and protect the spacing so three days never pass without a session.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

In 2003, two doctors stood next to each other for a DEXA scan. One was Chittaranjan Yajnik, an Indian endocrinologist. The other was John Yudkin, a British physician. They had identical BMIs of 22.3. By every screening number a doctor would normally use, they were the same body.

Yajnik’s body fat was 21.2%. Yudkin’s was 9.1%.

They published a single page in The Lancet and called it the YY Paradox (Lancet, 2004). The point was small and devastating. Same weight, same height, same BMI. Two completely different bodies. The fat that drives heart attacks and diabetes was sitting where the scale and the tape measure couldn’t see it. (Twice the body fat at the same weight. The scale couldn’t tell you which body was which.)

The lesson is bigger than DEXA. Weight is one number, and possibly the least interesting one on your body. The four numbers that predict how you’ll age are hidden underneath it.

TL;DR

• Your weight is the sum of four things that age you very differently: fat mass, lean mass, visceral fat, and bone density. Each is independently predictive.

• Lean mass predicts falls, frailty, and all-cause mortality after 60.

• Visceral fat drives heart disease, diabetes, and fatty liver.

• Body fat percentage catches the “thin outside, fat inside” trap.

• Bone density triggers the only hard clinical decision in the set: treat for osteoporosis or not.

• South Asians carry more body fat and less limb muscle at the same BMI than every other ethnic group studied. The scale is more misleading for us, not less.

• The gold-standard tool that reads all four in one ten-minute scan: DEXA. $150-300, radiation about a day of background.

Weight is one metric. These four matter more.

Body weight is a sum. It is the only health number most people track because it is the only one a bathroom scale can produce. But the same 180 lbs can be a lean, strong body composition with low visceral fat, or a sarcopenic body with a fatty liver and an elevated fracture risk. The scale cannot tell you which one you have.

The four numbers it hides (fat mass, lean mass, visceral fat, and bone density) each track a different way the body fails over time. Each moves in response to different inputs. Each matters at a different age. Tracking your weight without knowing what’s underneath it is like tracking your portfolio’s total dollar value without knowing what’s in it. The number tells you what you have. It does not tell you what is working.

Lean mass is the strongest mortality signal in older adults. A meta-analysis of prospective cohort studies found that adults over 65 with low skeletal muscle mass had a 56% higher risk of all-cause mortality than those with normal muscle (PLOS One, 2023; systematic review and meta-analysis). The functional version is what most people feel before any lab confirms it: less strength, slower stairs, more falls, longer recovery from anything that puts you in a hospital bed. South Asians start with less to spare. Compared with European men of similar weight, height, and age, Asian Indians have significantly less skeletal muscle in their limbs along with more central fat (British Journal of Nutrition, 2009). What moves it: protein at 0.75 to 1g per pound of lean body mass, two to three resistance training sessions a week, enough sleep to recover, and optimized hormones. After 40, you are either building lean mass or losing it. There is no holding it steady by accident.

If you’re on a GLP-1, this matters more, not less.
In the SURMOUNT-1 DXA substudy, about 25% of every pound lost on tirzepatide (Mounjaro, Zepbound) was lean tissue (Diabetes, Obesity & Metabolism, 2025; n=160). A real-world analysis of nearly 8,000 patients found tirzepatide drove greater lean mass loss than semaglutide (Ozempic, Wegovy) in routine care (medRxiv, 2026; n=7,965). Without a DEXA, the scale tells you the drug is working. The DEXA tells you whether your protein and training are keeping the right tissue.

Visceral fat, the deep abdominal fat that wraps your liver and pancreas, drives heart attacks, type 2 diabetes, and fatty liver progression independent of BMI or total body fat. On your DEXA report it shows up in pounds or square centimeters; above 100 cm² (or about 2 lbs) is the threshold most clinicians flag as elevated cardiometabolic risk (Hologic DXA reference; Nicklas et al., Obesity Research, 2004). What moves it: cut refined carbs and ultra-processed food, add fiber and lean protein, layer in Zone 2 cardio two to three times a week, and fix sleep. We’ll cover the drivers and treatments as their own topic in the next issue.

Body fat percentage catches the “thin outside, fat inside” pattern the YY Paradox first showed. In a recent NHANES analysis of US adults aged 20 to 49, body fat above 27% in men or 44% in women carried a 78% higher all-cause mortality risk over 15 years and more than three times the risk of heart disease mortality, while BMI lost statistical significance after adjustment (Annals of Family Medicine, 2025). The BMI says you are fine. The body fat number says whether that is true. Healthy ranges run roughly 10-20% for men and 20-30% for women, though the right target depends on age and goals. What moves it: a sustained caloric deficit if you are cutting, paired with the resistance training that keeps lean mass on. The same sleep and stress hygiene that moves every other metric on this list. There is no shortcut that doesn’t show up as muscle loss on the next scan.

Bone mineral density is the only DEXA number that drives a hard clinical decision: whether to treat for osteoporosis. Roughly 1 in 3 women and 1 in 5 men over age 50 will experience an osteoporotic fracture in their remaining lifetime (International Osteoporosis Foundation; Kanis et al., Osteoporos Int, 2000). The T-score on your report tells you where you stand. Above -1 is normal. Between -1 and -2.5 is osteopenia. Below -2.5 is osteoporosis. South Asian women have lower bone mineral density on standard DEXA scans, and the T-score reference data was built on Caucasian populations. A “normal” T-score can still underestimate fracture risk for South Asian women (Zengin et al., Frontiers in Endocrinology, 2015; review). What moves it: adequate calcium, magnesium, and vitamin D; weight-bearing and resistance training; no smoking and alcohol kept low. Hormone therapy is FDA-approved for prevention of bone loss, and for women within 10 years of menopause and under 60, the Menopause Society endorses it for bone protection, not just symptoms (NAMS, Menopause, 2022).

DEXA: The gold standard for measurement

There are other options. Bathroom scales with bioelectrical impedance (BIA) are convenient and cheap, but their fat-percentage readings drift several points with hydration and time of day. Skinfold calipers depend entirely on the technician. The Bod Pod (air displacement) is accurate but rare and pricey. MRI is the most accurate way to measure visceral fat and ectopic fat inside the liver, but it runs $1,500 or more per scan and isn’t ordered for routine body composition tracking.

A DEXA scan shines two low-energy X-ray beams through your body. Bone, muscle, and fat absorb them differently, and the machine reconstructs how much of each you carry, and where. Ten minutes on a table. $150-300 in most US cities. For body composition tracking, it is the most practical tool we have. DEXA-measured visceral fat correlates with MRI at r=0.88 to 0.94 across validation studies (Obesity, 2022). MRI is more accurate at the extremes; for everyday tracking, it is overkill at ten times the price.

DEXA has real limits. It tends to overestimate visceral fat at very high volumes (above ~600 cm³) compared with MRI (Obesity, 2022), and different machines disagree by 1-2%, so the trend on the same machine matters more than any single absolute number. It also doesn’t see ectopic fat inside the liver, pancreas, or heart muscle, which is what an MRI-PDFF or cardiac MRI is for. The least significant change for visceral fat on a GE CoreScan is about 130 grams (Obesity, 2019). Anything smaller than that is measurement noise. Chase trends across two or three scans, not single readings.

What about radiation?

A whole-body DEXA delivers about 2-10 microsieverts (µSv) per scan, depending on the machine. Most centers come in around 5 µSv (COMARE 18th Report, UK Department of Health, 2019; review of DXA dose data). That is roughly one day of natural background radiation in the US, where the average person absorbs about 3,000 µSv per year from soil, cosmic rays, and radon at home.

Exposure Effective dose Whole-body DEXA ~5 µSv Transcontinental flight (NY to LA) 30-40 µSv Chest X-ray 100 µSv Mammogram 280-400 µSv Cardiac CT (calcium scoring) 1,700-3,000 µSv Abdominal CT 8,000-10,000 µSv US background, full year 3,000 µSv

A DEXA is roughly 20 times lower than a chest X-ray and 1,000 times lower than an abdominal CT. The COMARE review estimated that a single DXA scan raises lifetime cancer risk by less than 1 in 1,000,000. Annual scans for body composition tracking, even sustained over decades, sit well inside the 20,000 µSv/year diagnostic exposure limit clinicians use as a guardrail. Pregnancy is the only common reason to defer.

How I use DEXA

I get a baseline DEXA every year. If I’m running a serious cut, a bulk, or starting a new medication that affects body composition, I add scans at the start and end of that window. The baseline is what makes the after number meaningful. Same scanning center every time. Machine-to-machine drift of 1-2% can swamp the real signal.

The first line I check on a new scan is lean mass. If lean dropped faster than fat in proportion, something in the plan is wrong. Protein is too low, the training stimulus isn’t enough, sleep is off, or the intervention is moving the wrong tissue. The report is most useful as an early warning, not a victory lap.

The clearest example I have is a six-month body recomposition window when I was using Tirzepatide. Between July 2024 and January 2025, I went from 244 to 226.5 lbs. The bathroom scale showed minus 17.5 pounds. The DEXA showed something different: fat down from 95 lbs to 69.5 lbs and lean mass up from 141 lbs to 149.8 lbs. Visceral fat down from 5.5 lbs to 3.89. Without the DEXA, the scale would have represented the change inaccurately. I have also had the opposite outcome, where I lost more lean mass than fat. DEXA caught it at the next scan, and I changed the plan before the next cycle.

Myth: “If the scale is going down, I am succeeding.”

The Myth: Weight loss is the goal. If the number on the scale is moving in the right direction, the plan is working.

The Evidence: Body weight is a sum. It hides what changed. In the SURMOUNT-1 DXA substudy of adults on tirzepatide, 25% of every pound lost was lean tissue (Diabetes, Obesity & Metabolism, 2025; n=160). The real-world digital phenotyping study found tirzepatide patients lost more lean mass than semaglutide patients, and the patients who lost the most lean mass also reported the most fatigue and reduced exercise tolerance (medRxiv, 2026; n=7,965). A separate review of older adults losing weight without structured strength training documented similar proportional muscle loss across multiple interventions (Mayo Clinic Proceedings, 2019; meta-analysis). The same 15 pounds on the scale can be a healthy recomposition, a clean fat loss, or a quiet muscle loss the reader will feel five years from now.

The Verdict: The weighing scale measures gravity. The DEXA measures what changed in the body. Especially after 40, and especially as a South Asian starting with less muscle to spare, the detailed numbers matter a lot.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

4-6 weeks in, the CGM wasn’t telling me anything new. I developed a sense of how my body responded to every food I regularly eat. I kept wearing it past the 4-6 week, and it unexpectedly became an accountability tool.

TL;DR

• The first 4-6 weeks of CGM data is genuinely revelatory. Individual glycemic responses vary by 5x for identical foods, and population nutrition charts can’t predict yours

• After that window, the new-data well runs dry. You’ve mapped your staples and the CGM stops surprising you

• The medical critics are right that CGM data is not a substitute for HbA1c in non-diabetics

• Beyond the learning window, the CGM is useful for two things: behavioral accountability (the Hawthorne effect) and early warning on metabolic shifts or drug/supplement side effects

Month one: your personal nutrition map

The first month is where the value lives. The Stanford “glucotypes” study put CGMs on 57 mostly healthy people and found 80% spiked to prediabetic or diabetic glucose levels after eating cornflakes with milk (PLOS Biology, 2018; observational, n=57). The lead researcher’s takeaway: “There are lots of folks running around with their glucose levels spiking, and they don’t even know it.”

The Weizmann Institute tracked 800 participants wearing CGMs for a week, measuring 46,898 meals, and found 5x variation in glycemic response to identical foods between individuals (Zeevi et al., Cell, 2015; observational cohort, n=800). A 2025 study confirmed it: each person has a carbohydrate “fingerprint,” and rice-spikers are more likely to be Asian, which, given my own CGM history, was not exactly a surprise (Wu et al., Nature Medicine, 2025; crossover, n=55).

That variance is the whole reason population nutrition charts fail you. A “low-GI” food is an average across hundreds of people. You might respond very differently to the same food compared to your friend.

So what do you do? Here’s how to spend four weeks with the device.

Week 1: log meals and watch the spikes. What you eat, when you eat it, what happens to your glucose. The spikes you see in your usual pattern are your baseline.

Weeks 2-4: experiment. Some worth testing:

• Eating vegetables and protein first, vs starting with rice or roti

• Building out the plate (more fiber, vegetables, healthy fats, protein), vs starchy carbs alone

• Different starch sources (rice, roti, oats, potato) to see which one spikes you most

• A 15-20 min walk after a meal, vs sitting on the couch

• 10 squats every hour during a long sitting day, vs sitting straight through

• If you are tracking your sleep, see how a good night’s sleep impacts the next day

• A teaspoon of psyllium husk (Isabgol) 30 minutes before the meal

You’ll see which foods consistently spike you, which conditions blunt the spike or cause it to come down quickly, and which interventions actually moved the needle. That’s your personal guide.

After month one: the diagnostic limits

The educational value has a natural end. Once you’ve cycled through your regular foods, the spikes stop being news. You learn how you respond to dal-chawal, your usual breakfast, your post-workout snack. After that, the CGM is showing you variations on patterns you already know.

This is also where the medical critics are right. A 2025 Mass General Brigham study of 972 adults found that CGM metrics correlated well with HbA1c in people with type 2 diabetes, weakly in prediabetes, and showed essentially no correlation in people with normal blood sugar (Diabetes Technology and Therapeutics, 2025; observational). University of Bath researchers found CGMs overestimated “time above range” by nearly 400% compared to gold-standard finger-prick tests in healthy adults, and misclassified low-GI fruits as medium or high-GI (American Journal of Clinical Nutrition, Feb 2025). The lead author explained the gap simply: CGMs measure glucose in the interstitial fluid around your cells, not in your blood, which introduces lag and inaccuracy at the edges.

The critics are evaluating the CGM as a diagnostic tool. As a diagnostic tool for healthy adults, it doesn’t really work, and the error rate is part of why. CGMs are accurate enough for someone whose glucose swings from 80 to 250, because the device’s error band is small relative to the signal. In a healthy adult whose glucose lives between 80 and 140, that same error band is large enough to make a normal post-meal reading look like a spike, or vice versa. The 400% overestimate is what happens when you push an interstitial-fluid sensor into a range it wasn’t optimized for. For most people, the natural endpoint is somewhere around four to six weeks. They’ve learned what they came to learn, and the device has nothing new to add.

Keep it on for accountability

For some people, including me, there’s a second reason to keep wearing it. It’s the Hawthorne effect: when people know they’re being observed, their behavior shifts toward what they think they should be doing. The data doesn’t have to be new for the device to keep working on you.

When my CGM is on, I eat fewer carbs at night, I snack less, I make better choices because I know the data is there, recording everything. When I take it off, I drift. I’ve tested this multiple times over five years. The pattern is consistent. I hate seeing big spikes on the graph, which makes me eat differently.

The same dynamic shows up in daily weighing. A study of 9,768 smart scale users found that only daily weighing was associated with weight loss across all BMI groups (JMIR, 2021; observational cohort, n=9,768). When participants stopped weighing for 30 or more days, they gained an average of 0.85 kg. The mechanism is identical: awareness triggers self-evaluation, which triggers corrective action.

If that kind of accountability is helpful for you, the CGM can be a very useful tool for blood sugar control. If it isn’t, or if it nudges you toward food anxiety instead of better choices, it’s the wrong tool.

Keep it on for early warning

There’s a third reason some people keep wearing one. When I started rosuvastatin (Crestor), I noticed within two weeks that my glucose spikes were getting larger and taking longer to come back down. Statins are known to increase insulin resistance by approximately 24%, with the pancreas trying to compensate until the system tips toward type 2 diabetes (ATVB, 2021). But HbA1c takes two to three months to reflect those changes. My CGM showed the shift in 14 days. I caught a real drug side effect months before my next blood draw would have flagged it.

Separately, I was waking up at 5am every morning for weeks without understanding why. When I looked at my overnight glucose traces, the pattern was obvious. My glucose was dipping in the early morning hours, triggering a cortisol and adrenaline response that pulled me awake (peak window for nocturnal hypoglycemia is between 04:00 and 07:30). I started having a piece of fruit with cottage cheese about an hour before bed, and the early wake-ups stopped within days.

Neither of those would have shown up on a standard lab panel. Postprandial glucose abnormalities can appear years before fasting glucose and HbA1c start to drift, and CGM is the practical way to see them (Journal of Multidisciplinary Healthcare, 2024; review). For South Asians, who develop metabolic dysfunction at lower thresholds, catching these shifts early may matter even more. A 2025 study found CGM metrics could differentiate prediabetes from normoglycemia in Asian individuals even when fasting glucose appeared normal (Chandran et al., Journal of Diabetes Science and Technology, 2025; observational).

Myth: “If you don’t have diabetes, a CGM is a waste of money”

The Myth: CGMs are medical devices for managing diabetes. If your blood sugar is normal, wearing one is expensive wellness theater, paying $100 a month to watch numbers that don’t mean anything clinically.

The Evidence: The belief persists because CGMs were marketed as optimization tools by wellness companies, and endocrinologists rightly pushed back on that framing. The critics are right about the diagnostic claim. But they’re evaluating only one of the device’s possible uses. The first four to six weeks of CGM data are genuinely educational for most people, regardless of diabetes status, because individual glycemic variability is large enough that you cannot predict your response to a food from population averages. After the learning window, the CGM becomes optional. For some people, behavioral accountability (the same psychology that makes daily weigh-ins work for weight loss) makes the cost worth it. For others, early detection of drug side effects or metabolic shifts justifies keeping it on. Most people probably should wear one for a month, learn what they came to learn, and take it off.

The Verdict: The first four weeks are worth it for almost anyone. After that, only if the accountability or early warning value is something you specifically need.

The CGM is a small example of a bigger pattern. Personal data lets you stop arguing with averages. Once you know how your body responds, you stop needing somebody else’s chart. The first month of wearing one is the lesson. What you do after is yours to decide.

How to Try One

You don’t need a prescription anymore. Dexcom Stelo (~$99/month for two 15-day sensors) and Abbott Lingo (~$49 for two weeks, $89 for four) are both OTC and ship directly from the company. Pick one, wear it for a month, run your food experiments. If you’re already on medications that affect glucose (statins, steroids, beta-blockers) or you have a family history of diabetes, loop your doctor in so the data lives in your file. Otherwise, the barrier to entry is lower than it’s ever been. OTC CGMs are slightly less accurate than prescription versions, but unless you have a metabolic diagnosis or risk flag, most doctors won’t prescribe one and insurance won’t cover it. For a one-month learning experiment, OTC is the path.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

Every South Asian family has some version of this argument. The aunties swear by ghee (loudly). The cardiologists say cut it. The internet is split between “ancestral superfood” and “arterial poison.”

I am a Punjabi and grew up in Delhi, with ghee in our kitchen. In the parathas, on the dal, stirred into the rice. It was love on the plate. So when I started looking at the evidence, I wanted a real answer.

The largest meta-analysis ever done on ghee and heart disease screened 919 articles and pulled 18 studies covering nearly 20,000 people, roughly 4,500 of them from the Indian subcontinent (Progress in Nutrition, 2025; systematic review and meta-analysis, n=19,948). The result was wide enough to mean “slightly protective” or “slightly harmful” or anything in between.

After all that data, the most honest thing science can say is: it depends. And that “it depends” pointed me somewhere useful.

TL;DR

• Whether ghee raises your numbers depends on your phenotype, your plate, and your labs, not on population averages.

• The 2025 meta-analysis (n=19,948) found no significant heart disease risk from ghee on average. Averages mask wide individual variance.

• A 2022 crossover RCT (n=30) showed ghee raised ApoB by +9 mg/dL vs olive oil on average. Your response could be very different.

• Hyperproducers vs hyperabsorbers explains the variance. A sterol panel tells you which you are.

919 articles later, science shrugged

That wide range is a shrug. The cleanest direct comparison sharpens the picture a little. A crossover RCT compared ghee against olive oil for four weeks each and found ghee raised ApoB by +9 mg/dL and non-HDL cholesterol by +20.5 mg/dL, both statistically significant (British Journal of Nutrition, 2022; crossover RCT, n=30).

ApoB is the atherogenic particle count, the metric that predicts cardiovascular events better than LDL alone. The signal is real, but n=30 over four weeks is an average across a tiny group, not a prescription for any one person.

One reason the population data is so messy: decades of Indian epidemiological research is confounded by vanaspati. Starting in the 1960s, hydrogenated vegetable oil containing up to 40% trans fats was marketed across India as “vegetable ghee.” Many studies on “ghee consumption” never distinguished desi ghee from vanaspati. The most atherogenic dietary fat in history was literally sold under the same name, and in many Indian households, nobody knew the difference. When someone cites Indian studies on ghee, the first question should be: which ghee?

Three levers, all individualized

Three things decide whether ghee is a problem for you: the plate, your biology, and your numbers.

The plate. One teaspoon of ghee finishing a lentil dal is a completely different metabolic event than three tablespoons soaking a paratha with sweet chai. Same fat, different context. The fiber, the dose, the carbohydrate environment around the fat, that’s where the metabolic impact lives.

The numbers add up fast: one tablespoon of ghee delivers roughly 10 grams of saturated fat, half to three-quarters of most recommended daily limits. Traditional Indian cooking often uses two to four tablespoons per dish, and most people aren’t measuring. And many South Asians who “cut ghee” replace it with more roti and more rice, swapping saturated fat for refined carbohydrate, which doesn’t help and often hurts. The plate around the fat matters as much as the fat itself.

Your biology. The biggest split inside individual biology is hyperproducers vs hyperabsorbers. Hyperproducers make most of their cholesterol in the liver, largely independent of what they eat. Hyperabsorbers pull most of theirs from food and bile through the gut. The same plate of ghee hits these two phenotypes very differently.

It’s the same reason statins (which block production) are transformative for some patients and underwhelming for others, while ezetimibe and psyllium (which block absorption) are dramatic for one group and barely move the needle for another. Psyllium is what most South Asian households know as Isabgol. Dietary saturated fat is on the same axis. Whether ghee raises your ApoB depends on which side of this split you’re on.

Genetic variants (FTO, PPARG, TCF7L2, ApoE), family history, and current metabolic state layer on top of the phenotype. None of them are fully captured by population averages.

Your numbers. Genetics points the direction, family history sets the risk context, and the labs give the specifics. If you want to go deeper than ApoB, a sterol panel measures synthesis markers (lathosterol, desmosterol) and absorption markers (campesterol, sitosterol), and tells you which side of the hyperproducer / hyperabsorber line you’re on.

My genetic testing flagged a saturated fat intolerance cluster across four gene variants (FTO, PPARG, TCF7L2, PER2), and my father had a heart attack at 69. We swapped ghee for olive oil as our everyday cooking fat. Ghee still shows up for flavor when it matters, but the daily default changed. The swap is a part of a broader lipid management program.

The trial data won’t resolve this for you

The ghee vs olive oil RCT is small (n=30) and short (4 weeks). The meta-analysis is large but inconclusive. Better trials would help, but even a perfect population trial wouldn’t tell you what to do, because population averages hide the responder / non-responder split that’s the whole story.

Your own data already can. Get a baseline lipid panel with ApoB, change one variable, retest in 6-8 weeks. If you want to know the underlying mechanism, get a sterol panel and see which side of the hyperproducer / hyperabsorber line you’re on.

Myth: “Our grandparents ate ghee every day and lived to 90”

The Myth: Ghee has been a staple of Indian cooking for thousands of years. Our grandparents ate it daily and didn’t get heart disease. Modern dietary advice is Western nutrition science ignoring ancestral wisdom.

The Evidence: The ancestral argument cherry-picks one ingredient from a lifestyle package you no longer live. Your grandmother also walked miles a day, ate fewer total calories, consumed far less refined carbohydrate, and had no vanaspati in her kitchen. Ghee existed inside a completely different plate and a completely different life. You can’t isolate one variable and credit it with the outcome.

There’s also survivorship bias at work. You’re remembering the grandparents who lived to 90. Take Neil’s family: all four of his grandparents had heart attacks. India has one of the highest age-standardized CVD mortality rates in the world.

And the study this argument usually rests on (Malhotra’s 1967 observation of 1.15 million Indian railway workers, British Heart Journal, 1967; ecological study) has no individual-level data. Northern India ate more ghee and had less heart disease, but physical activity, urbanization, diet composition, and vanaspati exposure all differed between regions. It cannot establish that ghee prevented anything.

The cultural verdict on ghee depends entirely on where you’re standing. The grandmother who fed it to you, the mom who cut it out, the wife who calls it a superfood. Same ingredient, three opposite stories, all with conviction and none with your data.

The Verdict: It’s Complicated. Both “ghee is poison” and “ghee is fine” are blanket statements that ignore the variance the science actually shows. Your grandparents’ ghee existed inside a lifestyle you don’t live anymore, and “grandma ate it” tells you nothing about your biology. The honest answer for you lives in your labs, not in the culture war.

The ghee question doesn’t have a universal answer, and that’s what makes it useful. What it does to your numbers depends on your plate, your phenotype, and your labs. The cultural debate becomes a personal one, and the personal one is the one you can act on. Between the auntie, the cardiologist, and the American wife who calls it a superfood, there’s another voice worth listening to: your own data.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

I stopped rosuvastatin because the metabolic cost was too high for someone already managing metabolic syndrome (covered in Part 1). Other people stop for muscle pain. Many never start because they’ve heard statins cause dementia, a worry the data doesn’t support: meta-analyses of randomized trials show no impact on cognition or dementia risk, and large observational studies have found a small reduction in dementia risk among statin users (Ott et al., Journal of General Internal Medicine, 2015; meta-analysis of RCTs).

For a while, I assumed my options were limited. They weren’t. Four FDA-approved drug classes now have cardiovascular outcome data, meaning they’ve been proven to prevent heart attacks and strokes, not just lower a number on a lab report. Most people who stop or avoid statins never hear about them. And one humble soluble fiber has real LDL-lowering evidence that slots in alongside these drug classes.

If you stopped a statin and haven’t replaced it with anything, you’re not alone. The “statins or nothing” framing is years out of date. Here’s what actually exists.

The Bottom Line

• Four non-statin drug classes now have cardiovascular outcome data: ezetimibe, PCSK9 inhibitors, bempedoic acid, and (partially) inclisiran

• Moderate statin + ezetimibe achieves similar LDL targets with 18% less diabetes risk than high-intensity statin alone

• PCSK9 inhibitors are now as low as $25/month with insurance

• Psyllium husk is the one non-drug adjunct with real LDL and ApoB data. Covered in Part 2 of this series.

• Every year of delayed LDL lowering compounds plaque buildup. Mendelian data shows roughly 3x more benefit from lifelong versus late-life lowering.

A $10 generic that made outcome trials boring

Ezetimibe (sold as Zetia) is the most accessible non-statin option and probably the most underappreciated. It blocks cholesterol absorption in the intestine through a different pathway than statins, so the effects are additive when combined.

IMPROVE-IT (NEJM, 2015; RCT, n=18,144) tested ezetimibe added to simvastatin in post-heart-attack patients. The result: a 6.4% relative reduction in the primary composite endpoint. That’s a real but modest benefit, which is why it never made headlines. The diabetes profile is what makes ezetimibe interesting for this audience. A pooled analysis of 4 cohort studies found that moderate statin + ezetimibe achieved similar LDL targets with 18% less new-onset diabetes than high-intensity statin monotherapy. Same destination, different metabolic cost.

Generic ezetimibe costs roughly $10/month. For someone concerned about statin side effects or diabetes risk, moderate statin + ezetimibe may be the smarter combination. Ezetimibe alone (about 18-20% LDL reduction as monotherapy) is also a real option for someone fully off statins, though the combination evidence is stronger. Worth asking about.

The injectable that costs less than you think

PCSK9 inhibitors (the antibody versions are sold as Repatha and Praluent) are the most powerful LDL-lowering tools available. They’re monoclonal antibodies that block PCSK9, a protein that normally breaks down LDL receptors on liver cells. Blocking PCSK9 means your liver clears more LDL from your blood, and the therapy is delivered as a subcutaneous injection every two to four weeks depending on dose.

The outcome data is strong. FOURIER (NEJM, 2017; RCT, n=27,564) showed a 15% reduction in the primary cardiovascular endpoint with evolocumab. ODYSSEY Outcomes (NEJM, 2018; RCT, n=18,924) showed a 15% reduction with alirocumab. Combined, that’s over 46,000 patients in landmark trials with hard endpoints. No diabetes signal in either trial, despite massive LDL lowering.

The cost story has changed dramatically. Repatha is available at $239/month through the AmgenNow direct-to-patient program (launched October 2025), roughly 60% below the list price. With some insurance coverage, co-pays can be as low as $25/month. The “$14,000/year” number that scared everyone off was 2015 pricing.

One caveat worth noting: inclisiran works differently from PCSK9 antibodies. It’s a small interfering RNA that silences PCSK9 production at the genetic level, which is why a single injection lasts about six months. The technology is genuinely novel and the side effect profile, in theory, should be minimal. But the cardiovascular outcome data isn’t in yet. ORION-4 (n=~15,000) is the phase 3 outcome trial that will answer whether LDL lowering translates into event reduction. Until that reports, inclisiran belongs in a different evidence category than ezetimibe, PCSK9 antibodies, or bempedoic acid.

Built for people who can’t take statins

Bempedoic acid (sold as Nexletol) is the newest option on this list and the most relevant for statin-intolerant patients. It works on the same cholesterol pathway as statins but upstream, and here’s the key difference: it’s a prodrug that activates only in the liver, not in skeletal muscle. No muscle activation means no myalgia, which is the side effect that sends most people to the supplement aisle in the first place.

CLEAR Outcomes (NEJM, 2023; RCT, n=13,970) was the first trial to show cardiovascular benefit in statin-intolerant patients specifically. The result: 13% reduction in major cardiovascular events. The diabetes profile was favorable, trending protective with no HbA1c or glucose increase. In a prespecified diabetes subgroup analysis, the benefit was even stronger: a 2.4% absolute risk reduction in major cardiovascular events (Lancet Diabetes & Endocrinology, 2024).

For someone who genuinely can’t tolerate any statin, bempedoic acid + ezetimibe is the combination with the strongest evidence base, achieving roughly 38% LDL reduction combined.

One practical wrinkle worth knowing before the appointment: with how insurance approvals work today, it’s now often easier to get a PCSK9 inhibitor covered than bempedoic acid, even for statin-intolerant patients. Coverage favors the drug class with longer and stronger outcome data, even when the newer option fits the clinical profile better. A small access reality that can flip the order of what you actually try first.

The soluble fiber that still earns its place

Before leaving the toolkit, the one non-drug adjunct: psyllium husk, what most South Asian households know as Isabgol. Part 2 of this series covered the evidence in detail (the meta-analysis, the mechanism, the dose, the cost, the cross-link to my deeper write-up). It belongs in any non-statin conversation, and it’s been part of my own regimen.

Every year you wait, your arteries remember

The urgency behind all of this comes from one concept: cumulative LDL exposure. Atherosclerosis is driven by the total LDL-years your arteries have experienced, not your LDL level today.

The data on this is striking. People born with genetic variants that naturally lower LDL by roughly 40 mg/dL from birth experience a 54.5% reduction in coronary heart disease risk (Ference et al., JACC, 2012; meta-analysis, n=312,321). That’s roughly three times the benefit seen when a statin delivers the same 40 mg/dL reduction starting in middle age, according to the same analysis. The European Atherosclerosis Society consensus confirms the effect is both causal and cumulative.

For South Asians, with coronary disease onset roughly 10 years earlier than European populations, the window for early treatment is narrower and the cost of delay is steeper. The question has shifted from “statin or nothing” to how soon, by how much, and with which tool.

How low is low enough?

Whatever you settle on, the next question is how aggressively to push the numbers. The 2019 ESC/EAS guidelines, still current as of the 2025 focused update, call for LDL below 55 mg/dL for very-high-risk patients (existing cardiovascular disease, diabetes with target organ damage, familial hypercholesterolemia) and below 40 mg/dL for those with recurrent events on maximum therapy (Mach et al., European Heart Journal, 2019; consensus guidelines). The American ACC/AHA guidelines sit at LDL below 70 mg/dL for similar profiles, moving in the same direction.

ApoB tracks alongside. The widely cited targets: below 80 mg/dL for high-risk and below 65 mg/dL for very-high-risk patients.

For South Asians, where 10-year ASCVD risk is systematically underestimated by standard calculators, the case for hitting the lower end is stronger. Practically: if you have a strong family history of early heart attack, stroke, or diabetes, treat yourself as “high risk” or “very high risk” even without a prior cardiac event of your own. The biology of cumulative LDL exposure is the same; the starting risk isn’t.

What these trials can’t tell us yet

The PCSK9 inhibitor and bempedoic acid trials enrolled predominantly white populations. South Asians are underrepresented in all four major outcome trials covered here. The direction of evidence is strong, but ethnic-specific benefit-risk ratios are estimated, not proven. Inclisiran’s outcome data (ORION-4) is still pending. And the optimal sequencing of non-statin therapies (which to try first, which combinations work best for which risk profiles) is still being worked out in clinical practice. The evidence base is expanding faster than the guidelines.

Myth: “If I can’t tolerate statins, there’s nothing else I can do”

The Myth: Statins are the only proven way to lower LDL and prevent cardiovascular events. If you get side effects, your choice is suffering through them or going unprotected.

The Evidence: The belief persists because of three things: guideline lag, provider inertia, and patient messaging that stopped at “take your statin.” Most primary care physicians trained when statins were the only game with outcome data, and guidelines take years to catch up with trial results. The cost narrative reinforced it, because PCSK9 inhibitors really were unaffordable in 2015. The evidence covered above tells a different story: three non-statin drug classes now have independent cardiovascular outcome data, costs have dropped dramatically, and one class was specifically designed for statin-intolerant patients. The toolkit changed. The conversation didn’t.

The Verdict: Statin intolerance is a real problem, but “nothing else works” hasn’t been true for years. The toolkit expanded while the narrative stayed frozen.

One Thing I Changed

For the past couple of years, Zetia plus psyllium husk did the heavy lifting. My ApoB went from a peak of 103 to 83 on that combination, without a statin in the picture. Recently I added Repatha, a PCSK9 inhibitor. Early days. I’m watching the numbers settle before deciding what the long-term regimen looks like.

The pattern across the full three-post arc is simple. Stopping one tool is not the same as stopping the work. The toolkit is bigger than most doctors will tell you, and the decisions are granular.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

The second came when I looked up what was actually in the bottle. The active ingredient in red yeast rice is monacolin K. Monacolin K is chemically identical to lovastatin, a prescription statin. The “natural alternative” IS the drug, sold without the dosing consistency, purity testing, or regulatory oversight that the prescription version requires. When researchers tested twelve red yeast rice products, they found citrinin, a toxic mycotoxin linked to kidney damage, in four of them (Gordon et al., Archives of Internal Medicine, 2010; laboratory analysis).

That sent me down a rabbit hole on every supplement marketed as a statin replacement, and I applied the same question from Part 1: has this been proven to prevent heart attacks, or does it just move a number on a lab report? For almost all of them, the answer is “the second.” For one, the answer is different. Let me show you both.

The Bottom Line

• No “natural statin alternative” has a single large RCT showing it prevents heart attacks or strokes

• Red yeast rice IS lovastatin (a prescription statin) in unregulated form, with toxic contamination in 4 of 12 products tested

• Niacin was tested in 25,673 patients and caused more harm than benefit. Two large trials ended the enthusiasm.

• The FDA identified 776 supplements containing unapproved pharmaceutical ingredients from 2007-2016

• Psyllium is the one exception: real LDL and ApoB data, boring mechanism, part of my regimen. Covered in a dedicated section below.

If your red yeast rice “works,” you’re taking an unregulated statin

The paradox sharpens when you follow it to its conclusion: if your RYR product has enough monacolin K to lower your LDL, you’re taking a statin with no quality control over dose, purity, or contamination. If it doesn’t contain enough, it’s doing nothing.

The chemical identity isn’t a coincidence. The first statin was discovered in 1973, when a Japanese biochemist named Akira Endo isolated it from a Penicillium mold growing on rice at a Kyoto grain shop. Red yeast rice is rice fermented on a different fungus that makes monacolin K, a molecule structurally identical to lovastatin. Both came from fungi making compounds that block the same enzyme. What separates them is regulation: dose, purity, and outcome data.

The FDA recognized this two decades ago, ruling Cholestin (a popular RYR supplement) an unapproved drug in 1998 and warning consumers about unauthorized lovastatin in RYR products in 2007. The monacolin K content varies wildly across brands, which means your dose changes every time you open a new bottle. That citrinin contamination in 4 of 12 products isn’t a theoretical concern either. Citrinin is a mycotoxin linked to kidney damage.

The clinical evidence behind RYR is thin. The studies are small, short, and none have measured cardiovascular outcomes like heart attacks or strokes. Compare that to the statin outcome data from Part 1: over 100,000 patients across decades of randomized trials.

Here’s the practical absurdity: generic lovastatin costs $4-10/month with a prescription. It comes with regulated dosing, known purity, and decades of safety monitoring. The RYR supplement that contains the same molecule often costs more, with none of those guarantees.

Niacin had its chance. Two large trials ended it.

Niacin is the cautionary tale for this entire conversation. For years it was a mainstream cholesterol therapy, prescribed to raise HDL. Unlike most supplements on this list, it got tested in large, rigorous trials, and failed both.

AIM-HIGH (NEJM, 2011; RCT, n=3,414) tested extended-release niacin added to statin therapy. No reduction in cardiovascular events. The trial was stopped early for futility. HPS2-THRIVE (NEJM, 2014; n=25,673) was even more definitive: niacin plus laropiprant added to statin therapy showed no cardiovascular benefit and caused more side effects, including infections, bleeding, and new-onset diabetes.

Two trials, nearly 30,000 combined patients. Niacin improves the HDL number on your lab report. It does not prevent heart attacks.

Four more supplements, zero cardiovascular outcome data

Bergamot doesn’t work in rigorous trials. The 2024 placebo-controlled RCT (MDPI Foods, 2024; n=110) found no improvement on LDL or oxidative stress markers, despite earlier excitement from small Italian studies. No cardiovascular outcome data exists.

Plant sterols are the strangest case. They produce a real 8-10% LDL reduction, but Mendelian randomization studies have linked elevated circulating plant sterols with cardiovascular risk (Nutrients, 2023). Lowering LDL through this mechanism may not deliver the benefit the LDL drop implies, and no outcomes trial has tested whether it does.

CoQ10 is the most common “take this with your statin” recommendation, and the evidence doesn’t support it. The premise is logical (statins block CoQ10, so supplementing should reduce muscle pain), but meta-analyses of randomized trials show it doesn’t help (JAHA, 2018; n=1,776). The deeper finding from one NIH-funded trial: only 36% of patients with self-reported statin myalgia developed symptoms during blinded testing. The other 64% were reacting to the idea of the drug.

Over-the-counter fish oil isn’t the same as Vascepa. The STRENGTH trial (JAMA, 2020; RCT, n=13,078) of EPA+DHA combo found no cardiovascular benefit. Vascepa is a prescription purified EPA with outcome data from REDUCE-IT, but it’s a regulated drug, not a supplement.

Psyllium is the exception

Before closing the door on supplements entirely, one deserves a section of its own. Psyllium husk, what most South Asian households know as Isabgol, is the one supplement in this category with real LDL and ApoB data.

A meta-analysis of 28 trials found that 10.2g daily reduced LDL by an average of 13 mg/dL and also reduced ApoB (Jovanovski et al., American Journal of Clinical Nutrition, 2018; n=1,924). The mechanism is physical: psyllium forms a viscous gel in your gut that binds bile acids. Your liver, now short on bile acids, pulls cholesterol out of circulation to make more. The result is a reliable LDL drop that compounds when paired with other lipid-lowering therapy.

This is category-different from the supplement graveyard. It’s fiber: cheap, boring, and effective. It’s been part of my own regimen for over a year. Practical basics: 10g daily, mixed in a full glass of water, before meals. For the full protocol, brand selection, and the synergy with Zetia, see my deeper write-up on psyllium.

776 supplements, and nobody had to prove they worked first

The Dietary Supplement Health and Education Act of 1994 makes all of this possible. Under DSHEA, supplements don’t have to prove they’re effective or safe before being sold. The FDA can only act after harm has already been demonstrated.

The consequences are measurable. The FDA identified 776 supplements containing unapproved pharmaceutical ingredients between 2007 and 2016 (JAMA Network Open, 2018; regulatory analysis). Roughly 20% of those contained multiple unapproved drugs. A separate analysis across 13 countries found 14.8% of non-hormonal supplements contained undeclared anabolic steroids. Independent testing by ConsumerLab has found roughly one in three multivitamins fail quality standards.

I wouldn’t run a business without audited financial data. At some point I started asking myself whether I was applying the same standard to my cardiovascular health.

What absence of evidence actually means

To be fair, absence of evidence is not evidence of absence. Large cardiovascular outcome trials cost hundreds of millions of dollars, and without patent protection, no supplement manufacturer has the incentive to run one. That’s a real limitation of the system, not a mark against any individual product.

But the limitation doesn’t change the evidence standard. If you’re choosing between a treatment proven to prevent heart attacks in tens of thousands of patients and one that hasn’t been tested at that level, the gap is real regardless of why it exists.

Myth: “Natural supplements are safer than prescription drugs”

The Myth: Natural alternatives to statins are safer because they come from nature, have fewer side effects, and aren’t made by pharmaceutical companies.

The Evidence: Red yeast rice IS a pharmaceutical (lovastatin) sold without pharmaceutical quality control, with toxic citrinin contamination in 4 of 12 products tested (Archives of Internal Medicine, 2010). Niacin, when tested in a large trial (HPS2-THRIVE, n=25,673), caused more infections, bleeding, and new-onset diabetes than placebo. The FDA found 776 supplements with undisclosed pharmaceutical ingredients (JAMA Network Open, 2018). Meanwhile, generic statins ($4-10/month) have decades of safety data in over 100,000 patients, known side effect profiles, and regulatory oversight requiring ongoing adverse event reporting.

The Verdict: “Natural” is a marketing category, not a safety category. Unregulated supplements have less safety data, less quality control, and in some cases more dangerous contamination than the prescription drugs they claim to replace.

One Thing I Changed

The RYR experiment ended my curiosity about cholesterol supplements. Psyllium earned its place in my regimen; the rest went back to the shelf. When I read a supplement label now, I read it for what’s missing as much as what’s listed: the outcome data, the dose consistency, the contamination testing. That’s usually where the story is.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

Then I looked at the actual data and realized I was asking the wrong question. “Do statins cause diabetes?” sounds like it has a yes or no answer. It doesn’t. The useful questions are different, and they changed how I think about this decision and about health decisions in general.

The Bottom Line

• “Do statins cause diabetes?” is the wrong question. Three better ones: what’s the tradeoff, which statin, and what’s the real alternative?

• The tradeoff: for every 1 diabetes case, 5.4 cardiovascular events prevented. 62% of new cases occur in people already near the diabetes threshold.

• Not all statins are equal: pitavastatin is glucose-neutral, atorvastatin carries the highest risk

• No supplement (red yeast rice, berberine, bergamot, niacin) has cardiovascular outcome data

For every diabetes case, five cardiovascular events don’t happen

The fear is real. Statins do increase diabetes risk. The largest meta-analysis on this, covering 13 trials and 91,140 patients, found a 9% relative increase in new-onset diabetes (Sattar et al., Lancet, 2010; meta-analysis, n=91,140). Over 4 years, for every 255 people treated, 1 additional diabetes case. But for every roughly 40 mg/dL drop in LDL, those same 255 people experienced 5.4 fewer major coronary events (heart attacks and coronary deaths). The ratio is 5:1 in favor of treatment, and it gets wider once you add in strokes and revascularizations.

The randomized evidence tells the same story. JUPITER, a trial that prospectively tracked diabetes alongside cardiovascular outcomes, randomized 17,603 people without diabetes to rosuvastatin 20mg or placebo (Ridker et al., Lancet, 2012; RCT, n=17,603). Rosuvastatin produced 270 new diabetes cases vs. 216 on placebo, a 25% relative increase. But the risk concentrated in people with baseline metabolic syndrome, impaired fasting glucose, BMI of 30 or higher, or HbA1c above 6%. In those without any of those risk factors, diabetes incidence was unchanged.

The 2024 CTT update made the picture even clearer. Using individual participant data from large, double-blind RCTs, it found that 62% of new diabetes cases occurred in people already in the top quarter of baseline glucose (CTT Collaboration, Lancet Diabetes & Endocrinology, 2024). The HbA1c increase was 0.06-0.08%, which is clinically trivial. The people developing diabetes on statins were already approaching the threshold, and the statin pushed them over a line they were walking toward. “Statins reveal diabetes” is closer to the truth than “statins cause diabetes.”

Here’s what makes this personal for anyone reading this newsletter: the people most at risk for statin-induced diabetes (those with prediabetes, metabolic syndrome, visceral obesity) are the same people who benefit most from statins, because their baseline cardiovascular risk is highest.

For South Asians specifically, the Pooled Cohort Equation underestimates our risk by more than 2-fold (predicted 4.8% vs. actual 6.8% 10-year ASCVD rate; Patel et al., Circulation, 2021; UK Biobank, n=8,124 South Asian within a cohort of 458,000). We develop coronary artery disease roughly 10 years earlier than white populations (and if you’re South Asian reading this, that math probably applies to your family too). The 5:1 ratio for a general population is probably more favorable for us, though we lack the ethnic-specific trial data to prove it.

Any time someone tells you a medication “causes” something, ask what’s on the other side. The scary number without the benefit number is half a story.

Seven drugs, not one

Most people treat “statin” as a single thing. It’s seven different drugs with seven different metabolic profiles.

The diabetes risk varies meaningfully between them. At one end: pitavastatin, which most people haven’t heard of. A distributed network analysis across 10 real-world databases compared pitavastatin to atorvastatin and rosuvastatin and found pitavastatin had a 28% lower risk of new-onset diabetes (Seo et al., Cardiovascular Diabetology, 2022; n=87,734 across 10 databases). Glucose-neutral in practice, effective LDL lowering, available generically.

At the other end, atorvastatin and rosuvastatin carry the highest diabetes risk among commonly prescribed statins. The same Seo analysis found the incidence rate was consistently higher on both than on pitavastatin across all 10 databases.

Dose matters too. High-intensity statins carry a 36% proportional increase in diabetes risk, while low-to-moderate intensity statins carry only 10% (CTT Collaboration, Lancet Diabetes & Endocrinology, 2024). This is a lever you can pull. A moderate-intensity statin combined with ezetimibe achieves similar LDL targets with 18% less diabetes risk than a high-intensity statin alone (pooled meta-analysis of 4 cohort studies). Same destination, different metabolic cost.

Which one? At what dose? Combined with what? Those are the questions that matter when a statin is on the table.

Your supplement hasn’t been tested against a heart attack

The instinct to try something natural first is understandable. I’ve had that conversation in my own family. But when someone chooses berberine over a statin, or red yeast rice over rosuvastatin, they’re making an evidence decision whether they know it or not. The question is: what standard of proof am I requiring before I bet my cardiovascular health on it?

Statins have cardiovascular outcome data in over 100,000 patients across decades of randomized trials. No supplement, not red yeast rice, not berberine, not bergamot, not niacin, has a single large RCT showing it prevents heart attacks or strokes. Some move a number on a lab report. None have been proven to prevent the event that number is supposed to predict.

What the statin-type ranking can’t tell us yet

The head-to-head diabetes risk comparison across statins comes primarily from observational data and network analyses, not from randomized trials designed to compare one statin against another on metabolic outcomes. The pitavastatin data is encouraging but based on smaller populations than the atorvastatin and rosuvastatin evidence. And South Asians are underrepresented in nearly all major statin trials, which means the ethnic-specific tradeoff ratio (likely more favorable than 5:1) is estimated, not proven. The direction of the evidence is clear. The precision will sharpen as more data comes in.

Myth: “Statins cause diabetes, so you’re trading one disease for another”

The Myth: Taking a statin means trading heart protection for a new metabolic problem. The diabetes risk cancels out the cardiovascular benefit.

The Evidence: The “trade” framing assumes the two risks are equivalent. They aren’t. The cardiovascular events statins prevent (heart attacks, strokes, cardiovascular death) are acute and often irreversible. The diabetes cases statins contribute to are concentrated in people who were already metabolically close to the threshold, with HbA1c increases of 0.06-0.08%, manageable with monitoring and lifestyle. One side of the trade is catastrophic. The other is a lab value that crossed a line it was approaching anyway. The 5:1 ratio covered earlier quantifies this: the benefit outweighs the risk by a wide margin, and the risk itself is not the diabetes most people picture when they hear the word.

The Verdict: The risk is real, but calling it a “trade” overstates the diabetes side and understates the cardiovascular side. The better question: which statin, at what dose.

One Thing I Changed

I stopped my rosuvastatin after it raised my fasting insulin and A1c. I did not stop lowering LDL. Zetia plus psyllium husk carried the work for over a year, dropping my ApoB from a peak of 103 to 83 and LDL from 127 to 74. I just switched to Repatha, a PCSK9 inhibitor. Early days. I’m waiting for the numbers to settle before deciding what the long-term regimen looks like.

The framing shift is what stuck with me. Yes-or-no rarely fits a health decision. The questions that do, and that I now bring to every one: what’s the tradeoff, which intervention, and what’s the real alternative.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

When I calculated my HOMA-IR, a single score that quantifies insulin resistance, it came out to 5.54. Anything above 2.5 means insulin resistant. I was more than double that threshold, and no doctor flagged it.

That number connected directly to the fatty liver disease I’d had since age thirteen. Insulin resistance was driving the fat accumulation the whole time, and fasting insulin would have shown it years earlier than anything else in my labs. In our diagnostics post, we listed fasting insulin as one of the core markers. This is why.

The Bottom Line

• Fasting insulin rises 10-13 years before glucose or HbA1c becomes abnormal (Lancet, 2009; Whitehall II, n=6,538)

• 75% of glucose-normal people have abnormal insulin patterns (Kraft database, n=14,384)

• Track four numbers together: fasting insulin, fasting glucose, HbA1c, and HOMA-IR. Any one alone can miss what the others catch. HOMA-IR = (insulin x glucose) / 405. Above 2.5 = insulin resistant.

• A fasting insulin test costs $20-30. Most doctors don’t order it. Ask.

Your annual physical checks the last rung, not the first

Metabolic dysfunction works like a ladder. Insulin rises first, then triglycerides climb, HDL drops, and your liver starts storing fat. Years later, glucose and HbA1c finally move. Your annual physical checks the last rung. Fasting insulin checks the first.

The Whitehall II study tracked over 6,500 people for 13 years and found that future diabetics already had significantly lower insulin sensitivity 13 years before diagnosis (Lancet, 2009; prospective cohort, n=6,538). Glucose only spiked in the final 2-3 years. The body compensates for insulin resistance by pumping out more insulin, keeping glucose in range, until it can’t. Standard screening catches the final stage: when compensation fails and glucose finally spikes.

That’s where HOMA-IR comes in. It turns two numbers from a standard blood draw into one insulin resistance score: below 1.0 is optimal, above 2.5 is insulin resistant, and above 5.0 is severe. The formula is simple: (fasting insulin x fasting glucose) / 405. You can calculate it yourself from any lab that includes both numbers.

Why doesn’t your doctor order fasting insulin? Because it’s not in the ADA screening guidelines. The entire diabetes screening paradigm is built around glucose, which is the last thing to break, not the first. The test exists and costs $20-30. The guidelines just haven’t caught up.

The “normal” range is lying to you

The reference range on your lab report makes this worse. Most labs call anything between 2 and 18.4 µIU/mL normal, but a fasting insulin above 9.0 already identifies prediabetes in 80% of affected patients (Endocrine Practice, 2009; cross-sectional, n=2,000+). At 14.3 µIU/mL, my lab report said normal. I was already well into insulin resistance territory by functional standards.

For South Asians, the standard ranges are even more misleading. HOMA-IR values run 67% above European averages (Diabetologia, 2013; Ghouri et al., age- and BMI-matched, n=186). When both groups gained the same weight, South Asians lost insulin sensitivity at roughly five times the rate: a 38% decrease compared to 7% in white Europeans (Nature Metabolism, 2024; GlasVEGAS controlled feeding study, n=35). The margin for early detection is thinner, the consequences arrive sooner, and the standard cutoffs were built on a different population.

I’m obese, and my glucose said I was fine

I’ve been obese since childhood. I was diagnosed with fatty liver disease at thirteen. By any common-sense measure, something is clearly off metabolically. But look at the glucose and A1c columns below.

Date Insulin (µIU/mL) Glucose (mg/dL) HbA1c (%) HOMA-IR Feb 2023 14.3 - 5.3 - Sep 2023 13.1 - 5.2 - Dec 2023 26.7 84 5.6 5.54 Oct 2024 16.9 82 5.4 3.42 Dec 2024 13.6 92 5.2 3.09 Sep 2025 14.3 89 - 3.14 Jan 2026 8.2 78 5.2 1.58

Glucose never left the normal range. HbA1c barely moved. If you only looked at the standard panel, you’d say I was metabolically healthy. Meanwhile, fasting insulin was screaming. At its peak, my HOMA-IR hit 5.54, more than double the threshold for insulin resistance, and not a single doctor flagged it.

The numbers that caught the problem were the ones nobody ordered: fasting insulin and HOMA-IR. That’s how the test is designed.

What HOMA-IR can and can’t tell you

HOMA-IR is a fasting-only measure. It tells you nothing about what happens after a meal. Someone can have a normal HOMA-IR and still have terrible post-meal glucose spikes, which is what a CGM can reveal. The two work together: HOMA-IR catches the baseline, a CGM catches the response.

One caveat on the numbers themselves: if you get your insulin tested at two different labs, you might get two different results. The tests aren’t perfectly standardized yet. That’s fine. What matters is the trend over time at the same lab, not any single number. Pick a lab and stick with it.

An elevated HOMA-IR is a conversation with your doctor, not a self-diagnosis. It’s a screening tool that opens a door. I’m watching whether mine holds below 2.0 as I continue to lose weight.

Myth: “My glucose is normal, so I’m metabolically healthy”

The Myth: Your annual physical checks fasting glucose or HbA1c. If those numbers are normal, your metabolism is working fine. You’d know if something was wrong.

The Evidence: The Kraft database, validated across 14,384 patients with extended glucose-insulin tolerance tests, found that 75% of people with normal glucose had abnormal insulin patterns (Crofts et al., Diabetes Research and Clinical Practice, 2015; n=14,384). A study of 1,313 young adults in Mumbai found 30.5% had hyperinsulinemia with completely normal glucose, and among those with HbA1c below 5.7%, nearly 40% had stimulated insulin above 80 mIU/mL (Frontiers in Clinical Diabetes, 2023; n=1,313). Globally, 1 in 4 adults has insulin resistance (Frontiers in Endocrinology, 2025; meta-analysis, n=235,148). Most don’t know it because the test that catches it isn’t part of standard screening.

The Verdict: Normal glucose is the last domino to fall, not the first sign that everything is fine.

One Thing I Changed

I now test my fasting insulin in my quarterly draws, and track that along with HOMA-IR so I can get a better sense of my metabolic health over time.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

My ALT was always elevated, and every doctor shrugged and repeated the same line. Nobody mentioned my heart. Nobody mentioned that my liver was both a symptom and an accelerator of a metabolic problem building across multiple organ systems.

It turns out that fatty liver, now called MASLD (metabolic dysfunction-associated steatotic liver disease), sits at the center of a metabolic cascade that connects to cardiovascular disease, Type 2 diabetes, and kidney disease.

The Bottom Line

• Fatty liver is now MASLD: renamed in 2023 to reflect that it’s a metabolic condition, not a drinking problem

• 34.7% of South Asians have MASLD, with rates as high as 48% at BMI ≥25 (JGH, 2025; 62 studies, n=40,745)

• MASLD and heart disease share the same metabolic root: insulin resistance. The liver amplifies the signal, with significantly higher cardiovascular risk even after adjusting for other metabolic factors

• Standard liver enzymes (ALT/AST) are normal in ~25% of patients with significant fatty liver

One in three South Asians has it, and most don’t know

The largest meta-analysis of MASLD in South Asians pulled 62 studies covering 40,745 people and found a prevalence of 34.7% in the general population, rising to 60% among those with Type 2 diabetes (Journal of Gastroenterology and Hepatology, 2025; systematic review and meta-analysis, n=40,745). The mean BMI of those South Asian patients was only 24.8, barely above the normal range.

Globally, about 19% of normal-weight individuals have MASLD, and lean MASLD is more prevalent in Asia than anywhere else. A mean BMI of 24.8 means many of these patients wouldn’t trigger screening under standard Western guidelines. You can be at a weight most doctors would call healthy and still have fat accumulating in your liver.

The fibrosis data is what concerned me most. South Asian MASLD patients had advanced fibrosis (scarring) at a rate of 14%, roughly double the rates found in European and American cohorts (JGH, 2025). Lower BMI, higher fibrosis rates, worse outcomes at every stage.

The liver is a metabolic amplifier

For years I thought of my fatty liver as a liver problem. It’s actually a signal that the entire metabolic system is under stress, and the liver both reflects that stress and makes it worse.

The mechanism works through insulin resistance, the common root. When the liver accumulates fat, it becomes less responsive to insulin, which drives dangerous cholesterol patterns (more small dense LDL, higher triglycerides, lower HDL), chronic inflammation, and elevated blood sugar. Those same pathways drive cardiovascular disease and diabetes independently. The liver sits in the middle, amplifying all of it.

A nationwide Swedish cohort study (n=10,422) found that MASLD patients had 67% higher cardiovascular risk with noncirrhotic fibrosis and 115% higher with cirrhosis, even after accounting for traditional cardiometabolic risk factors. But most of the excess risk comes from the shared metabolic drivers (insulin resistance, inflammation, visceral fat), not from the liver itself. The liver is an accelerator, not the root cause.

The connection to diabetes runs both directions. Up to 65-70% of people with Type 2 diabetes have MASLD, and 15-30% of MASLD patients develop Type 2 diabetes within five years. In the South Asian meta-analysis, prevalence among T2DM patients was 60% (JGH, 2025). South Asians already have 1.5-2x higher insulin resistance at the same BMI as European populations.

Your ALT might be lying to you

Two problems make MASLD harder to catch than it should be. First, approximately 25% of MASLD patients have completely normal ALT and AST levels (BMC Gastroenterology, 2020; meta-analysis, 11 studies, n=4,084). Many patients are asymptomatic with liver function tests that look fine. A normal ALT does not mean a healthy liver.

Second, the diagnostic criteria were redesigned in 2023 when NAFLD was renamed MASLD. The new criteria use a BMI threshold of 23 for Asian populations, not 25 (which, if you’re South Asian, probably applies to you or someone in your family). That single number means South Asians qualify for screening at a lower weight than other groups, but most primary care doctors are still using the old cutoffs.

The practical starting point is a FIB-4 score, which uses four numbers from a standard blood panel (age, ALT, AST, platelets) to estimate fibrosis risk with no extra blood draw. If you’ve had a DEXA scan, check your visceral adipose tissue number: elevated visceral fat is strongly associated with liver fat, and you may already have that data sitting in a report. FibroScan is the next step if staging is needed, but start with what’s accessible.

28 years and counting

My ALT has been elevated for 28 years, and the January 2026 result of 84 U/L is still nearly double the upper limit.

But my DEXA scans tell a clearer story than my liver enzymes. Visceral fat dropped from 5.5 to 3.89 lbs over the past year, and my CCTA (coronary CT angiography) in November 2024 showed zero coronary atherosclerosis despite 28 years of NAFLD. DEXA and CCTA track progress that liver enzymes miss. If I had relied on ALT alone, I would have assumed nothing was changing.

I’m on tirzepatide, a GLP-1/GIP dual agonist. GLP-1 receptor agonists significantly reduce liver fat (some studies suggest reductions of up to almost 50%) (Wang et al., 2025; systematic review, 25 RCTs, n=2,600). For a condition that had zero approved treatments for decades, there are now three on the table (resmetirom, semaglutide for MASH, and tirzepatide showing the highest efficacy in late-stage trials).

What the data can’t tell us yet

The 2025 South Asian meta-analysis has very high heterogeneity, with prevalence ranging from 18.5% in rural settings to 47.1% in urban ones. Detection method matters enormously, and these numbers will sharpen as more studies use standardized MASLD criteria.

The Mendelian randomization evidence on CVD causality is nuanced: genetic variants that cause liver fat (PNPLA3, TM6SF2) show weak or no causal link to heart attacks, suggesting that the liver fat itself isn’t the direct CVD driver. The inflammatory pathways that come with it likely are. The clinical implication is that treating the metabolic dysfunction (insulin resistance, inflammation, visceral fat) matters more than treating the liver in isolation.

Most MASLD is driven by insulin resistance and caloric excess. Genetic factors like choline metabolism can play a role (they do in my case), but they’re one pathway among many.

Myth: “Fatty liver is caused by drinking too much”

The Myth: Fatty liver disease is a consequence of heavy alcohol use. If you don’t drink, or barely drink, you don’t need to worry about it.

The Evidence: MASLD (metabolic, non-alcohol-related fatty liver disease) now affects 38% of all adults globally, up from 25% two decades ago (Hepatology, 2024; 92-study meta-analysis). It is more common than alcohol-related liver disease worldwide. The primary driver is insulin resistance and metabolic dysfunction, not alcohol. The 2023 renaming from NAFLD to MASLD explicitly addressed this confusion: the old name implied that the default fatty liver was alcoholic. The new name drops the alcohol framing entirely and centers cardiometabolic risk. Many of the South Asian patients in the meta-analysis had a BMI that most doctors wouldn’t flag, and many have never had a drink.

The Verdict: Busted. MASLD is a metabolic condition, not a drinking problem. Insulin resistance, visceral fat, and metabolic dysfunction are the primary drivers.

One Thing I Changed

I added a gut-liver axis probiotic to my morning routine, targeting the connection between gut barrier function and liver inflammation (this took me a surprisingly deep research rabbit hole to land on a specific strain). Too early for results, but I’m tracking ALT at my next draw.

Twenty-eight years with this condition, and I’m only now understanding the full picture of what it connects to. The liver, the heart, blood sugar, kidney function: they’re all part of the same metabolic story.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

In our first post, we looked at why South Asians get heart disease a decade earlier than other populations, and why lifestyle alone doesn’t explain it. The natural next question is: what do you actually do about it?

It starts with knowing your numbers. Not the ones from a standard annual physical, which checks almost none of the markers that matter most for South Asians, but the right ones that help you establish a baseline and track what’s actually happening over time.

In this post, we go through the markers that matter, why they’re important, and how often to check them. The goal is to give you enough context to have an informed conversation with your doctor and to start building your own picture of your cardiometabolic health.

This is not medical advice. It’s what we think is worth testing, why each marker matters, and what we’ve learned from the data.

The blood work your annual physical skips

ApoB is the single most important lipid marker most doctors don’t order. It counts every atherogenic (plaque-building) particle in your blood. LDL cholesterol estimates how much cholesterol those particles carry, but ApoB tells you how many particles there are. Two people with the same LDL-C can have very different ApoB numbers and very different risk profiles. If we could only pick one lipid marker, it would be this one.

Lp(a) is a genetically determined cholesterol particle that you test once. You can’t diet or exercise it down. 25% of South Asians have elevated levels, and it accounts for 9.5% of heart attack risk in South Asians vs. about 4% in White populations (Circulation, 2019; INTERHEART Lp(a) sub-analysis, n=6,086). If your Lp(a) is high, it doesn’t mean you’re doomed. It means you manage everything else harder: ApoB, blood pressure, glucose. Knowing the number changes the strategy.

Fasting insulin is the one most doctors skip entirely. They’ll order fasting glucose, which is a snapshot of your blood sugar at one moment in time. The problem is that glucose can look normal for years while insulin climbs to compensate. By the time glucose is elevated, insulin resistance has been building for a while. Ask for both fasting glucose and fasting insulin every time. For South Asians, where prediabetes shows up nearly 8x more often by age 45 than in White peers (JAHA, 2026; MASALA/MESA, n=2,700), catching insulin resistance early is the difference between prevention and treatment.

HbA1c is a 90-day blood sugar average that smooths out the day-to-day noise. A single fasting glucose can be thrown off by what you ate the night before. HbA1c gives you the bigger picture. If you have a family history of Type 2 diabetes, this is the trend line that matters most quarter to quarter.

The full lipid panel (LDL, HDL, triglycerides) is where most doctors stop. For South Asians, the pattern to watch is high triglycerides combined with low HDL and small dense LDL particles, what cardiologists call atherogenic dyslipidemia. North Indians develop coronary artery disease at total cholesterol around 170 mg/dL (Indian Journal of Medical Research, 2007), a number most doctors would call normal. Triglycerides are also a useful cross-check against metabolic health: if they’re elevated, something is off with insulin signaling, regardless of what your glucose says.

hsCRP (high-sensitivity C-reactive protein) is a general inflammation marker. Inflammation is one of the drivers of atherosclerosis, and visceral fat is a common culprit. hsCRP moves with body composition: as visceral fat comes down, hsCRP tends to follow. It’s a signal that connects metabolic health to cardiovascular risk.

Homocysteine is an amino acid linked to cardiovascular risk and B vitamin metabolism. South Asians tend to run higher, partly due to higher prevalence of MTHFR variants and partly due to vegetarian diets that are low in B12. The fix is usually B vitamins, which is why we recommend also tracking B12, B6, and folate to make sure you’re actually replete. Elevated homocysteine is common and fixable, but you have to know it’s there.

Advanced lipid panel (NMR) is worth getting once a year. It breaks down particle count and size: LDL-P, small dense LDL, HDL particle number. Two people with the same LDL cholesterol can have very different risk profiles depending on how many particles they have and how big they are. That said, the markers that actually get treated are LDL and ApoB. If those are in a good place, the doctor isn’t going to treat the particle sub-fractions independently. The NMR gives context, but ApoB is where the clinical decisions happen.

The imaging that blood work can’t replace

Blood work tells you what’s in your bloodstream. Imaging tells you what it’s done to your arteries.

Coronary artery calcium (CAC) scan is a 10-minute CT scan, typically $75-150, and most places don’t require a referral. It measures calcified plaque in your arteries. MASALA data shows that 46% of South Asians under 50 already have detectable calcium (MASALA, PMC9830106). A score of 0 is reassuring. Anything above 0 tells your doctor how aggressive to be with prevention.

A Cleerly CCTA (coronary CT angiography) goes a step further, mapping both calcified and non-calcified plaque. It’s more expensive and involves contrast dye, so it’s not for everyone, but if you have significant family history, it gives a more complete picture than CAC alone.

DEXA scanning is one of the most underused tools in preventive health. Most people think of it as a bone density test, but a full-body DEXA shows body fat percentage, lean mass, and critically, visceral fat, the metabolically dangerous fat wrapped around your organs that drives inflammation and insulin resistance. A scale tells you your weight changed. A DEXA tells you what changed. At the same BMI, South Asians carry 30% more visceral fat than other populations (International Journal of Obesity, 2016; MASALA/MESA, n=906). The scale won’t catch that. A DEXA will.

Other relevant markers to consider

These aren’t the core cardiometabolic panel, but they each connect to the bigger picture.

Testosterone (free and total) matters more in your 40s than most men realize. Total testosterone is what most doctors check, but free testosterone is the fraction that’s actually bioavailable. A solid total number can mask low free testosterone. Low T affects body composition, insulin sensitivity, mood, and cardiovascular risk. Ask for both.

Estradiol and SHBG give context to the testosterone numbers. Estrogen matters in men too: too high or too low affects mood, body composition, and cardiovascular risk. SHBG (sex hormone binding globulin) binds testosterone and makes it unavailable, so it’s the context that makes the other hormone numbers make sense.

TSH, Free T3, Free T4 cover thyroid function. Most doctors only order TSH. Free T3 is the active hormone cells actually use, and TSH alone can miss subtle dysfunction. Ask for all three.

Ferritin and full iron panel (iron total, TIBC, transferrin, iron saturation) are important because ferritin depletes first, long before iron total or hemoglobin drop. Without the full panel, depletion doesn’t show up until it hits as fatigue and poor recovery.

Vitamin D deserves a mention because 84% of South Asian women are deficient. Darker skin, indoor lifestyles, and clothing coverage all contribute. Supplementation is common but the right dose varies by individual, so track quarterly until you find it.

Omega-3 Index and fatty acid panel measures whether supplementation is actually working. The target is an Omega-3 Index above 8%. The ratio between omega-6 and omega-3 tells you whether chronic inflammation is being fueled by diet.

TMAO (trimethylamine N-oxide) ties gut microbiome health to cardiovascular risk. It’s a newer marker and not many people track it, but given the cardiovascular focus of this newsletter, it’s worth knowing about.

eGFR is a kidney function check. Kidneys filter everything: supplements, medications, metabolic waste. If kidney function is declining, you want to know early.

What all and how often

Blood work

Imaging

Other markers to consider

How to get started

You don’t have to do all of this at once.

If you’re South Asian and haven’t done anything beyond a standard physical, start here: ApoB, Lp(a) (once), fasting insulin, HbA1c, and hsCRP. These five tests, added to your standard lipid panel, will tell you more about your cardiometabolic risk than most people learn in a decade of annual physicals. Most can be added to a standard blood draw by asking your doctor. If your doctor pushes back, services like Function Health, Marek Health, or a direct-to-consumer lab can run them without a referral.

From there, consider a CAC scan if you have family history or are over 35. Consider a DEXA if you want to know what’s happening under the surface with visceral fat and lean mass.

If you’re actively managing multiple risk factors, quarterly blood work timed with a doctor check-in is ideal. For someone in maintenance mode, twice a year or even annually is enough for most markers. Lp(a) is a one-time test. Calcium CT every 3-5 years. DEXA annually.

The point is to track the right things and establish a longitudinal record of your body so you can catch things in time.

Disclaimer: The information provided in this article is for educational and informational purposes only. It is not intended as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified healthcare providers with any questions you may have regarding a medical condition or wellness program. Reliance on any information provided in this article is solely at your own risk. The author and publisher of this article make no representations or warranties, express or implied, regarding the completeness, accuracy, reliability, suitability, or effectiveness of the information contained herein. The inclusion of specific products, services, or strategies in this article does not imply endorsement or recommendation. The author and publisher disclaim any liability for any adverse effects or consequences resulting from the use or application of the information presented. You are encouraged to consult with a qualified healthcare professional before making any changes to your diet, exercise routine, or lifestyle.

Then my family made it personal. My dad had a heart attack at 69. The angiography revealed 90% blockage in three arteries and 60-70% in three others. His cholesterol was “normal.” His blood sugar was fine. A few years later, my mom’s cousin collapsed at the breakfast table after a morning run, mid-50s, never drank, never smoked, healthy weight. He died before the ambulance arrived. A few months ago, another family member, seemingly in good health, passed away in his car while driving to the hospital.

One in ten heart attacks in South Asian men happens before age 40, roughly 2-3x the rate in Western Europeans (Lancet, 2004; INTERHEART, n=27,000+). Not 60 or even 50, but forty. Dr. Namratha Kandula, who leads the MASALA study on South Asian heart disease, put it simply: “Everyone knows somebody who has had a heart attack at a young age. It’s very common.”

I started paying attention.

TL;DR

• South Asians experience heart attacks 10 years earlier than other populations (median first heart attack: age 53 vs. 63)

• Risk factors like prediabetes appear 8x more often by age 45 in South Asian men, even with healthier lifestyles

• A CAC scan at 35-40 can show whether plaque has started building, years before symptoms

• Standard risk calculators weren’t built for you. Ask for A1c, ApoB, Lp(a), and imaging.

A decade earlier, and lifestyle alone doesn’t explain it

The INTERHEART study found that the median age of first heart attack in South Asians was 53 years, compared to 63 in Western Europe and China (Lancet, 2004; case-control, n=27,000+). A full decade earlier. And 9.7% of South Asian men had their first heart attack before age 40, roughly 2-3x the rate in Western Europeans.

The MASALA study tracked South Asian Americans head-to-head against other ethnic groups and found that despite better diet quality, lower alcohol consumption, and comparable exercise levels, South Asians still carried dramatically higher cardiovascular risk (JAHA, 2026; MASALA/MESA cohort comparison, n=2,700).

South Asian men at age 45 were nearly 8x more likely to have prediabetes than White peers (30.7% vs. 3.9%), with metabolic risk factors appearing at significantly younger ages. The lifestyle was fine, but the biology underneath it wasn’t.

The ruler was built for a different population

At the same BMI, South Asians carry significantly more fat in the places that matter most: visceral fat wrapped around organs, fat deposited inside liver tissue, and fat threaded between muscles (International Journal of Obesity, 2016; MASALA/MESA cross-sectional, n=906). You can look lean and still have a fatty liver driving insulin resistance, and standard BMI cutoffs won’t flag it.

Your cholesterol panel is probably incomplete too. South Asians tend toward a pattern of high triglycerides, low HDL, and small dense LDL particles, what cardiologists call atherogenic dyslipidemia. North Indians develop coronary artery disease at total cholesterol around 170 mg/dL (Indian Journal of Medical Research, 2007), a level most doctors would call normal.

Then there’s Lp(a), a genetically determined cholesterol particle that 25% of South Asians have at elevated levels. It accounts for 9.5% of heart attack risk in South Asians vs. about 4% in White populations (Circulation, 2019; INTERHEART Lp(a) sub-analysis, n=6,086 cases). You can’t diet or exercise it down. The only move is to test it once, know your number, and if it’s high, manage everything else harder.

My ApoB was 103 and nobody blinked

When I finally asked for ApoB, the number that counts every atherogenic particle in your blood, mine came back at 103 mg/dL. My LDL-C at the time was 127 mg/dL, technically flagged high but not a number most doctors would lose sleep over (I know, I was surprised too). Through Zetia, tirzepatide, and psyllium husk, my ApoB is now 83 mg/dL and my LDL-C is 74 mg/dL.

In November 2024, I got a Cleerly CCTA, a coronary CT angiogram that maps plaque in your arteries. It showed my arteries are clean, and my CAC score came back at 0. Reassuring, but not permanent. I’ll repeat it in 3-5 years.

The numbers being good now isn’t the point. Nobody would have caught the problem if I hadn’t asked for the right tests, and I only knew to ask because I went looking.

Good studies, done on different populations

INTERHEART is a case-control study, not a prospective trial, so the 53 vs. 63 comparison has design limitations. The “one in ten before age 40” data is from the global INTERHEART study and may not map perfectly onto diaspora South Asians, though the MASALA study shows similar patterns in South Asian Americans.

Even after accounting for ectopic fat, lipids, and Lp(a), researchers found these factors don’t completely explain the elevated coronary calcium in South Asians (JAHA, 2017; MASALA sub-study, n=803). Something else is contributing, and the theories are still being debated. One possibility: generations of calorie scarcity shaped a metabolism that stores energy efficiently, now mismatched with modern abundance. A 2024 study found that South Asian men who gained weight lost 38% of their insulin sensitivity vs. 7% for European men gaining the same amount (Nature Metabolism, 2024; controlled overfeeding study, n=35). Another thread points to epigenetic inheritance, though this evidence is suggestive rather than conclusive.

The data doesn’t wait for the explanation. Only 3.16% of participants across 310 cardiovascular clinical trials were South Asian (JACC: Asia, 2025; systematic review, n=1M+). The treatment guidelines your doctor follows were largely built on data from other populations.

Myth: “If I eat right and exercise, I should be fine”

The Myth: Heart disease is a lifestyle problem. Maintain a healthy diet, stay active, don’t smoke or drink, and your cardiovascular risk is low.

The Evidence: The MASALA study found that South Asians with better diets and lower alcohol intake still carried dramatically worse cardiovascular risk than other ethnic groups. Munaf Patel is the human version of that data point. He had his first heart attack at 39. He changed everything afterward: diet, exercise, stress management. Eight months later, he had a second heart attack. His father died of heart disease in his 40s (Chicago Tribune, 2019).

Lifestyle absolutely matters, and it’s the first thing any cardiologist will recommend. But for South Asians, lifestyle alone isn’t enough. Our biology works against us in ways that diet and exercise can’t fully offset: excess visceral and liver fat at the same BMI, Lp(a) levels you inherited and can’t change, and an atherogenic lipid pattern that a standard panel doesn’t even measure.

The Verdict: Lifestyle matters, but for South Asians, biology sets a different baseline.

There’s no shortage of health content out there: newsletters, podcasts, longevity influencers. But almost none of it is written for the South Asian body. The dietary advice assumes a Western diet. The risk calculators were calibrated on European populations. The clinical trials enrolled almost no South Asians. And most of our community has no idea how different our risk profile actually is.

I’m writing Desi Vitals with my dear friend Dr. Karan Bhalla. Karan is a board-certified interventional cardiologist in Houston, Texas. He trained at Seth G.S. Medical College in Mumbai, did his cardiology and interventional fellowships at UTHealth McGovern, and co-founded Orion Medical, where he’s been practicing for over fifteen years. He’s a Fellow of the American College of Cardiology, past president of the Harris County Medical Society, and faculty at McGovern Medical School. He sees South Asian patients with exactly these risk patterns every day. I bring the patient side, Karan brings the clinical side, and together we aim to do the work of reading the research, talking to the experts, and synthesizing it into clear, actionable information so you don’t have to spend hours sorting through scientific papers and podcasts to figure it out yourself.

Desi Vitals is the newsletter we wish had existed when we first started paying attention. Evidence-based South Asian cardiometabolic health, written for our community, by people who’ve lived it.