In this episode of The Feldman Protocol, how should researchers and clinicians weigh insulin resistance against lipid markers when assessing cardiovascular risk? Dave Feldman sits down with Sameer Dossani (PhD, anthropology; health coach specializing in autoimmune and gut conditions) to explore human evolutionary dietary history, the origins of metabolic disease across ancient cultures, the lean mass hypersponder phenotype and ketone production, challenges in science communication, and the role of AI in personalized health.

::: spoiler generated summary

Personal history and the turn toward nutrition (01:05-14:40)

• My health coaching centers on autoimmune disease, especially gut conditions; a recently completed anthropology PhD and personal history shape this approach.
• A South Asian Khoja Ismaili family history runs through Gujarat, Calcutta and modern Bangladesh; the 1971 war preceded the family's long residence in the US and a Washington, DC, childhood.
• Human origins are African, with migrations out of Africa tens of thousands of years ago; mixed South Asian ancestry and a trading family history illustrate human mobility.
• An unhappy elite-school experience preceded university in Canada, philosophy and religion studies, participation in punk and vegan communities, and activism against international economic institutions.
• Roughly two decades in international economics and NGOs, including ActionAid, involved work across several countries; the 2015 move to South Africa, layoffs, and family health problems led toward doctoral work and health coaching.
• Medical-library reading contradicted earlier beliefs about plant-based nutrition; repeated failure to correct medical and dietary advice became more troubling than isolated error.
• Entrenched training, institutional incentives, and inherited assumptions can perpetuate a hypothesis despite contrary data, without requiring a coordinated group of individuals.

Diabetes before modern dietary guidelines (14:40-25:05)

• Medical training depends heavily on memorization, while anthropology and philosophy of science add questions about the origins and historical development of medical ideas.
• Ancient Indian descriptions of madhumeha associated diabetes with sweet urine and carbohydrate restriction; Chinese medical advice attributed to Sun Simiao also included avoidance of carbohydrates and alcohol.
• Carbohydrate restriction has a long history in diabetes care; modern diabetes organizations have also supported carbohydrate-inclusive diets and invoked the brain's carbohydrate requirements.
• A childhood type 1 diabetes diagnosis in North Africa or West Asia involved immediate restriction of bread, an emphasis on meat and above-ground vegetables, and concurrent insulin therapy.
• Carbohydrate restriction for children with type 1 diabetes can provoke concerns over social inclusion and quality of life, alongside the practice of dosing insulin for carbohydrate-containing foods.
• Food pleasure, social convenience and disease prevention can pull in different directions; a culturally popular food can still aggravate an existing health condition.

Comfort foods, ancestral diets and evolutionary hypotheses (25:05-48:30)

• Concern about eating disorders can lead to broad acceptance of every food; differentiating highly processed food-like substances from traditional foods is distinct from viewing all restrictions as disordered.
• A doughnut is a modern comfort food, not an ancestral necessity; enjoyment, dietary balancing and family tradition do not establish a health benefit.
• Highly rewarding processed foods and widespread social participation normalize consumption, with historical parallels to the social acceptance of smoking.
• Blood-pressure concerns coexist with strong family and Bangladeshi sweet-food traditions; allulose and homemade lower-carbohydrate ice cream are practical ways to accommodate these habits.
• Human brain expansion and a smaller gut began roughly 1.3-1.5 million years ago; cooking and greater consumption of animal fat are two explanations, while securely dated fire use occurs later in this chronology.
• Evolutionary diets provide the starting hypothesis; modern hunter-gatherer dietary variety supports a generalist model, while a specialist model centers large, fatty prey.
• [1] A July archaeological analysis across the Americas yields estimates of roughly 90-99% or more of edible animal biomass from megafauna; mammoths and other large prey anchor the specialist-hunter model.
• Megafaunal extinction around 12,000 years ago may have forced dietary diversification; the specialist-hunter account remains a working hypothesis about prehistory.
• Human gut proportions, stomach acidity, forward-facing eyes and sweating are compatible with a meat-oriented evolutionary account; persistence hunting among modern San communities can also have ritual or tourist functions.
• The evolutionary loss of fur and reliance on sweating increased the importance of skin pigmentation; melanin offers solar protection, while lighter skin supports vitamin D production under weaker sunlight.

Famine cuisine, dietary dependence and communication (48:30-01:03:00)

• Stress encourages self-soothing through food and other comforts; older recipes differ from modern products, and a diet useful during famine need not optimize health during abundance.
• Repeated South Asian famines under British rule form part of the history of inherited susceptibilities and starch-centered food customs, complicating current dietary change.
• Sweet cravings and prior stress can persist during dietary change; coffee and lower-carbohydrate substitutes can displace other foods, while a Quest cookie and eggs differ nutritionally and emotionally.
• Public communication connecting metabolic health with the 2020 infectious-disease crisis encountered platform restrictions; earlier videos were deleted after channel growth stalled, and roughly one in nine podcast episodes required removal from YouTube.
• Metabolic health was a correlate of severity and outcome during the infectious-disease crisis; opportunities for preventive dialogue existed amid polarized public reactions.
• Scientific uncertainty requires precise language; the evolutionary dietary model is a revisable starting hypothesis, and public audiences can imitate incidental choices such as diet soda or snack foods.

Veganism, personal illness and LDL (01:03:00-01:22:50)

• Veganism began around age 18 within a Montreal punk community, followed by ethical, environmental and health commitments; eggs and dairy entered the diet later, with intermittent returns to vegan eating.
• Vitiligo stabilized during part of the vegan period, while asthma did not improve alongside smoking; a vegetarian pregnancy was complicated by gestational diabetes.
• By 2015-2016, a largely plant-based diet coincided with knee pain, reduced night vision, severe reactions to milk, very low vitamin D, HbA1c around 6.5 and weight gain; other family members also had health problems.
• South African interest in Banting and Tim Noakes revived an earlier conflict between carbohydrate-centered and dietary-fat-centered explanations of metabolic illness.
• [2] A Gerald Reaven study separated participants by measured insulin resistance; clinical events clustered in the most insulin-resistant group, with none in the least-resistant group, while LDL had little discriminatory value within that analysis.
• Initial changes included elimination of seed oils, a five-day water fast, carbohydrate restriction, sun exposure and greater intake of animal foods; that abrupt sequence is not suitable for unprepared imitation.
• Weight fell from roughly 98 to 78 kg, with improvements in vitamin D and other measures; LDL increased to roughly 350 mg/dL and later above 400, creating concern despite concurrent improvements.
• A CAC score of zero and improved well-being coexist with markedly elevated LDL; the preferred starting point is physiological adaptation and evidence of disease rather than an isolated laboratory value.
• LDL responses to ketogenic eating vary even within a family; substantial subjective gains can bias interpretation of a laboratory marker moving in an unfavorable direction.
• [3] The identical-twin vegan-versus-omnivorous trial used LDL as its main outcome and found a between-group difference of roughly 14 mg/dL, a modest absolute change given its prominence in public debate.
• Work on religious influence within dietetics associations offers one explanation for persistence of plant-based assumptions; much nutritional epidemiology has small effect estimates and possible confounding.

Epidemiology, anomalies and unreleased trial data (01:22:50-01:41:15)

• [4] Bradford Hill's 1965 causality criteria include strength of association; doubling the relative risk of a rare disease can yield a small absolute increase.
• Relative risks around 1.15 often underpin headlines about meat and cancer; observational associations are useful for generating hypotheses without automatically establishing causation.
• Nurses' Health Study and Physicians' Health Study dietary analyses use questionnaire-based dietary intake and face healthy-user confounding through smoking, alcohol, exercise and related behaviors.
• [5] A Harvard doctoral analysis found an inverse association between ice-cream consumption and cardiovascular outcomes among people with diabetes, complicating attempts to isolate benefits of yogurt.
• The ice-cream association could involve dairy fat, intake-recording behavior, confounding or chance; post hoc analyses can preserve a favored yogurt hypothesis without resolving the anomaly.
• [6][7] Chris Ramsden reanalyzed recovered Minnesota Coronary Experiment and Sydney Diet Heart Study data; an original investigator withheld findings because the results were unwelcome.
• Incomplete publication of unfavorable trial findings undermines trust in nutrition research; falsifiable hypotheses, attempts at disproof and publication of negative results are central to scientific practice.
• [8] Metformin's interference with resistance-training muscle gains preceded a reversal in Peter Attia's personal use; confidence in earlier pharmaceutical interventions warrants scrutiny alongside willingness to update them.
• LDL goals often sit below the levels common in adults, sometimes with recommendations for early lowering; LDL had been mildly elevated even during a vegan diet and could be lowered with oatmeal.
• A large breakfast of oats, berries and honey preceded disabling hunger during tennis and a second large meal; dietary LDL reduction can coexist with undesirable glucose and appetite effects in insulin resistance.
• The full effects of an intervention matter beyond one biomarker; better LDL values can coexist with worsening glycemic control and other health indicators.

Medical institutions and the KETO-CTA results (01:41:15-02:08:30)

• Professional guidelines carry institutional authority, while the term fad diet can dismiss patient-led dietary changes undertaken after unsatisfactory diabetes management.
• Ketosis probably occurred repeatedly during human history under limited food availability; modern highly processed diets are more novel than periodic ketosis.
• A physician initially concerned about LDL later adopted low-carbohydrate eating; other doctors in the practice also changed habits, illustrating shifts in clinical assumptions through experience.
• [9] The KETO baseline analysis compared carbohydrate-restricted people with high LDL and favorable HDL and triglycerides against matched Miami Heart participants; this phenotype differs from insulin-resistant atherogenic dyslipidemia.
• [10] The longitudinal KETO-CTA analysis found an association of baseline plaque with later progression but no within-cohort association of LDL or ApoB with progression; average LDL was roughly 254 mg/dL following years of elevation.
• Lean mass hyper-responders are increasingly recognized in clinical practice but remain heterogeneous, with some individuals developing plaque or faster progression.
• [11] PARADIGM linked annualized percent atheroma volume growth around or above 1% to higher event risk; average progression was about 0.93% in event cases and 0.32% without events.
• In KETO-CTA, 21% under QAngio and 29% under HeartFlow exceeded the 1% threshold, while median progression was about 0.23% and 0.27%; individual outcomes cannot be inferred from the cohort median.
• A model in which cumulative ApoB exposure alone predicts individual plaque progression is incompatible with the absence of a direct ApoB-progression association within this cohort; additional factors remain relevant.
• [12] Maasai observations in George Mann's research include low recognized coronary mortality alongside arterial plaques at autopsy; plaque burden and clinical events are distinct outcomes.
• Insulin resistance changes HDL, triglycerides and LDL particle patterns while influencing cardiovascular risk, creating confounding in simple LDL-outcome associations.
• Statin therapy has beneficial outcomes in some high-risk populations, and PCSK9 therapy has relevant data; statins also influence inflammation and nitric oxide pathways, complicating mechanistic attribution.
• Antibiotics can be worthwhile during a serious infection, while preventive medication and population-wide exposure involve tradeoffs, informed consent and patient choice.

Muscle, exercise and dietary self-assessment (02:08:30-02:23:20)

• Resistance training is central to preserving muscle in older adults; increased meat, reduced carbohydrate intake and improved insulin sensitivity can accompany better DEXA measurements in coaching experience.
• Protein intake near one gram per pound of ideal body weight and slow-to-failure resistance sessions lasting roughly 15-20 minutes twice weekly form Dave's muscle-building routine.
• An L1 compression fracture during weight use underscores injury concerns; machines and regular lower-intensity exercise address safety and mental well-being alongside muscular strength.
• A dietary trial can reveal previously normalized symptoms; knee pain, dandruff and longstanding bowel irregularity improved following dietary change.
• A strict 30-, 60- or 90-day ancestral-style dietary trial offers a way to observe changes in symptoms and well-being against habitual eating.

Fiber, bowel health and autoimmune gut disease (02:23:20-02:34:15)

• [13] A 2012 chronic-constipation intervention compared differing fiber intakes and found complete relief in the no-fiber subgroup, challenging the assumption that increased fiber always helps.
• A high-fat, hypercarnivorous pattern combines meat with additional fat to approximate the fat content attributed to large wild prey.
• Abrupt fasting and carbohydrate restriction can produce digestive distress; a gradual four- to six-week transition to higher fat intake can reduce diarrhea and bile-related symptoms.
• Major carbohydrate restriction changes electrolyte needs; insufficient sodium, potassium or magnesium can accompany fatigue, nausea and malaise.
• Inflammatory bowel disease, including Crohn's disease and ulcerative colitis, is a focus of dietary coaching; multiple sclerosis also has a body of dietary intervention research.
• Childhood vitiligo worsened during stress and later improved during a period of carnivore eating, with continued improvement during a high-fat, animal-dominant diet that includes vegetables and spices.
• During inflammatory bowel disease flares, puréed meat, soup and broth can reduce the burden of large solid meals; two or three smaller meals can be easier than one large meal.
• Large late meals can disrupt sleep; food intake ending at least two to three hours before bed, or three to four hours in Dave's routine, supports better nighttime rest.

AI, expertise and limits of diagnosis (02:34:15-02:46:30)

• AI-assisted symptom analysis can expand the diagnostic possibilities for rare diseases but can also produce alarming, unlikely explanations and unnecessary medical visits.
• Refsum syndrome in two brothers involved impaired handling of phytanic acid from foods including beef and dairy; earlier identification could reduce cumulative exposure and related visual damage.
• AI outputs depend on training data, clinical guidelines and developer constraints; sustained dialogue can expose neglected factors while also generating agreement with a user's assumptions.
• An anthropology inquiry involving nitrogen-isotope evidence for omnivores initially produced an incorrect exclusion of a paper; more specific bibliographic clues corrected the retrieval, demonstrating the importance of domain expertise.
• Continuous AI-assisted debate about carbohydrates does not itself change diet or health; action requires a workable decision despite uncertainty.

Lean mass hyper-responders, ketones and VLDL (02:46:30-02:52:45)

• [14] The lipid energy model connects high LDL in lean, carbohydrate-restricted individuals to fatty-acid mobilization, hepatic VLDL production and delivery of triglycerides to peripheral tissues.
• Hepatic fatty acids can enter ketone production or triglyceride-rich VLDL synthesis; a modest shift toward VLDL can coexist with high LDL and low circulating beta-hydroxybutyrate.
• A highly fat-adapted person can have beta-hydroxybutyrate near 0.2-0.5 mmol/L while oxidizing fatty acids; ketone concentration alone does not quantify overall fat use.
• A respiratory quotient near 0.70-0.75 would be consistent with substantial fat oxidation despite low ketones; this mechanism remains testable and not universal.

Research corrections, interpretation and changing medicine (02:52:45-03:04:19)

• [10] The April 7 longitudinal KETO-CTA paper required a significant correction and was subsequently retracted after anomalies in the Cleerly plaque data; acknowledgment and correction of errors are essential to scientific practice.
• Anonymized raw Cleerly data contained implausible patterns; repeated requests for paid, blinded quality control were declined, while QAngio and HeartFlow provided independent plaque analyses.
• The independent plaque analyses also linked baseline plaque with subsequent progression and found no detectable LDL/ApoB-progression association within the cohort; anonymized data were released for external analysis.
• Several prominent critics continued to dispute the work without accepting the Cleerly anomalies; Steven Quartz subsequently acknowledged a problem, making consistency in data scrutiny a central concern.
• Publication alone does not rapidly change entrenched clinical beliefs; institutional communication and repeated conversations have increased recognition of the lean mass hyper-responder phenotype.
• Premature cardiovascular disease in a wider family motivates concern about risk prioritization; severe hyperglycemia and high LDL can coexist, making metabolic dysfunction a major therapeutic priority.
• With HbA1c initially at 10 and elevated LDL, reducing HbA1c to five yields a different metabolic risk pattern from lowering LDL while HbA1c stays at 10; persistent severe hyperglycemia remains consequential.
• Clinical emphasis on LDL and ApoB can overshadow insulin resistance, whose association with disease is especially important outside special lipid disorders such as homozygous familial hypercholesterolemia.
• Recognition of lean mass hyper-responders is growing among clinicians, with research debates progressing at a pace faster than some academic expectations but slower than engineering expectations.

References

1. [00:40] Hemisphere-wide evidence of Early Paleoindian megaherbivore specialization — https://doi.org/10.1126/sciadv.aef9628
2. [01:12] Insulin resistance as a predictor of age-related diseases — https://doi.org/10.1210/jcem.86.8.7763
3. [01:20] Cardiometabolic Effects of Omnivorous vs Vegan Diets in Identical Twins: A Randomized Clinical Trial — https://doi.org/10.1001/jamanetworkopen.2023.44457
4. [01:23] The Environment and Disease: Association or Causation? — https://doi.org/10.1177/003591576505800503
5. [01:28] Dairy Products and Cardiometabolic Health Outcomes — https://dash.harvard.edu/bitstreams/90ef08c0-57ce-4ff9-b6ef-3dc7cdc4f7a3/download
6. [01:33] Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73) — https://doi.org/10.1136/bmj.i1246
7. [01:33] Use of dietary linoleic acid for secondary prevention of coronary heart disease and death: evaluation of recovered data from the Sydney Diet Heart Study and updated meta-analysis — https://doi.org/10.1136/bmj.e8707
8. [01:34] Metformin blunts muscle hypertrophy in response to progressive resistance exercise training in older adults: the MASTERS trial — https://doi.org/10.1111/acel.13039
9. [01:53] Carbohydrate Restriction-Induced Elevations in LDL-Cholesterol and Atherosclerosis: The KETO Trial — https://doi.org/10.1016/j.jacadv.2024.101109
10. [01:53] RETRACTED: Longitudinal Data From the KETO-CTA Study: Plaque Predicts Plaque, ApoB Does Not — https://doi.org/10.1016/j.jacadv.2025.101686
11. [01:55] Progression of whole-heart atherosclerosis by coronary CT and major adverse cardiovascular events — https://doi.org/10.1016/j.jcct.2020.12.007
12. [02:00] Atherosclerosis in the Masai — https://doi.org/10.1093/oxfordjournals.aje.a121365
13. [02:23] Stopping or reducing dietary fiber intake reduces constipation and its associated symptoms — https://doi.org/10.3748/wjg.v18.i33.4593
14. [02:47] The Lipid Energy Model: Reimagining Lipoprotein Function in the Context of Carbohydrate-Restricted Diets — https://doi.org/10.3390/metabo12050460 :::