5 November 2026 · 15 min read

The Island With No Recorded Strokes or Angina: What Kitava Can and Cannot Tell Us

Executive summary

  • The 1990 survey on Kitava, an island community in Papua New Guinea with traditional horticultural practices, revealed that none of the 213 adults interviewed reported experiencing stroke, sudden death, or angina. The research depended on interviews rather than imaging techniques or autopsies (Ref 1).
  • The diet primarily consisted of tubers, fruit, fish, and coconut, relying on local resources and containing a significant amount of carbohydrate, unaffected by Western dietary practices (Ref 1, Ref 2). The data do not address carnivore diets, as nobody on the island followed one.
  • The Kitavans maintained a slim physique, exhibited reduced diastolic blood pressure, had fasting insulin levels at half those observed in Swedish adults aged 50 to 74, and displayed low activity of PAI-1, a protein known to impede blood clot dissolution (Ref 3, Ref 4, Ref 5).
  • The population of Kitava does not exemplify individuals with elevated LDL levels. The cholesterol carried by their LDL was 10 to 30 per cent lower compared to Swedish reference populations among older adults (Ref 3).
  • The study observed that the Kitavan triglyceride and HDL levels were unfavourable, and several clotting factors exceeded those of Swedish participants (Ref 5, Ref 6).
  • The Tsimane people in Bolivia, whose diet consists of approximately 72 per cent carbohydrates, exhibited minimal coronary artery calcium on CT scans despite experiencing significant inflammation (Ref 7).
  • Consensus statements widely accepted in the mainstream and investigations using Mendelian randomisation both provide compelling evidence that LDL triggers cardiovascular disease in a dose-dependent manner (Ref 8, Ref 9).
  • Based on my current understanding, factors such as leanness, fasting insulin levels, and low diastolic blood pressure appear to account for the lack of cardiovascular disease observed on Kitava more effectively than LDL does (Ref 3, Ref 4). If I'm mistaken, I will revise my stance should the data change.

My initial stance requires disclosure. I operate a private clinic providing fee-based consultations on metabolic health, and I typically follow and advise an animal-based, carnivore dietary approach. The stance I hold is that LDL does not cause atherosclerosis or heart disease, with metabolic markers instead being much more significant in determining risk. This stance, rather than a discovery, influences how I interpret this research. In an effort to offset the bias, I have referenced the strongest version of the prevailing argument for LDL while also including the data that are not favourable to my perspective.

In 1990, a study was conducted on Kitava Island, part of the Trobriand group, where approximately 2,250 people maintained a traditional horticultural lifestyle (Ref 10). This is an observational report rather than an experimental study. The study does not demonstrate that saturated fat poses no risk or that coconut offers benefits, nor can it alone refute the accepted theories of heart disease. The study examines a basic interpretation of the diet-heart hypothesis, which states that "saturated fat raises LDL, LDL causes heart disease", but the findings do not align well with this straightforward model.

What was actually measured on Kitava

The researchers conducted semi-structured interviews involving a group of 213 adults ranging in age from 20 to 96 years old. According to the authors, "no case corresponding to stroke, sudden death or angina pectoris was described by the interviewed subjects" (Ref 1). The authors estimated that 125 islanders fell within the 60 to 96 age range, indicating the presence of a significant portion of individuals susceptible to these conditions (Ref 1).

The results are based on recollected incidents, not autopsies or scans, and a heart attack specifically is not included in the interview questions detailed in the abstract. Electrocardiograms (ECGs) at rest were recorded for 171 adults. The readings were not entirely normal, as 14 ECGs exhibited Minnesota-code abnormalities, despite the authors' conclusion that stroke and ischaemic heart disease "appear to be absent" (Ref 1). In this one-time visit to a group heavily affected by parasites and infections, it emerged that 76 per cent of men and 80 per cent of women above 20 were smokers, while every adult consumed betel nut (Ref 3, Ref 10).

The dietary staples and the carbohydrate question

The diet primarily consisted of cultivated tubers such as yam, sweet potato and taro, along with minor quantities of cassava, and was complemented by fruit, leaves, nuts like coconut, fish, maize and beans (Ref 10). The diet lacked dairy, refined fat, sugar and alcohol, with a low salt intake (Ref 2). The authors described the average physical activity level as moderate to active, at approximately 1.7 times the basal metabolic rate (Ref 10).

This was clearly a diet high in carbohydrates, just like the one consumed by the Tsimane, which will be addressed later (Ref 7). Carbohydrates are, in my view, not essential for anyone, even athletes, but the findings must be presented as they were reported. People who are lean and active and consume unrefined traditional carbohydrate can exhibit minimal recorded cardiovascular disease.

This does not support the idea that carbohydrate is necessary, beneficial, or safe for inactive individuals in modern food settings. The information gathered does not address the impact of a carnivore diet, as neither group followed one. Hypothetically speaking, factors such as leanness, regular physical activity, and avoiding refined sugar, refined fat, and ultra-processed foods might account for the lack of metabolic disease from carbohydrate intake in these scenarios.

The metabolic markers and the unflattering data

The Kitavans exhibited a notably lower body mass index, averaging 18.9, in contrast to Swedish reference groups (Ref 10). Blood pressure readings indicated substantially lower diastolic values that remained stable across ages, while systolic pressure increased steadily from age 50 onwards in both men and women (Ref 2, Ref 3).

Fasting insulin, a crucial metabolic marker in my assessment, was consistently reduced across all age groups in Kitava. Kitava's average fasting insulin levels for those aged 50 to 74 were "only 50%" of those in a different group of Swedes, with levels decreasing as people aged on the island but increasing in Swedes over 50 (Ref 4). According to the authors, elevated insulin levels indicate "insulin resistance," a condition I refer to as physiological insulin resistance, where cells resist glucose uptake; under this interpretation, the Kitavans exhibited a reduced level of this condition, likely associated with their lower body fat (Ref 4).

The data showed that 85 per cent of Kitavan men and all women had PAI-1 activity levels of 5 U/ml or lower, compared to 22 per cent of Swedish men and 14 per cent of Swedish women (Ref 5). The median level of C-reactive protein (CRP), which indicates inflammation, was 0.5 mg/L compared to 1.1 mg/L in Swedes, though there were notable caveats (Ref 10). The acceptance rate for blood sampling among islanders was 41 per cent, while the Swedish controls underwent sampling several years afterwards, and the effect size was minor. Removing 19 subjects with type 2 diabetes or hypertension erased the contrast for all CRP measurements, leaving it solely for levels under 10 mg/L (Ref 10).

The data do not universally praise the population. The study's authors referred to the Kitavan triglyceride and HDL profile as unfavourable (Ref 6). Triglycerides in 20 to 39-year-old Kitavans exceeded those in Swedes (Ref 3), but the researchers could not determine if this discrepancy arose from typical physiological responses or partly "the high intake of carbohydrates and saturated fat in Kitava" (Ref 6). In Kitavans, levels of factor VII, factor VIII, and von Willebrand factor were elevated, while fibrinogen was reduced by 10 per cent in males but increased by 25 per cent in females (Ref 5).

The description of Kitava as a population thriving despite elevated LDL levels due to coconut is inaccurate. In Kitavan males aged 40 and above, as well as females aged 60 and above, levels of total cholesterol, apolipoprotein B, and estimated LDL (measured by its cholesterol content) were 10 to 30 per cent lower compared to Swedish reference populations (Ref 3). Between 20 and 50, LDL increased in men (Ref 2). The authors suggested that having a lean physique and lower diastolic blood pressure "seem to offer the best explanations", while "the lower serum cholesterol may provide some additional benefit" (Ref 3). In my view, a variation of 10 to 30 per cent in a group with non-negligible LDL levels appears less likely than metabolic factors to explain the findings, although these data alone cannot confirm this.

Alternative explanations

Antitreponemal antibodies, potentially stemming from earlier yaws infections, were detected in 87 per cent of Kitavan blood samples, whereas the presence in healthy control groups ranged from 0 to 6 per cent (Ref 11). Cardiovascular disease's low prevalence on Kitava has been partly linked to natural IgM-antiphosphorylcholine antibodies, which may be triggered by treponemal infection (Ref 11). The notion continues to be a theoretical possibility that I do not discount.

Pacific islanders participating in the Tokelau Island Migrant Study exhibited increasing disease indicators after relocating to New Zealand. Male migrants to New Zealand exhibited adjusted systolic blood pressure levels that were 7.2 mmHg greater and diastolic levels that were 8.1 mmHg higher compared to non-migrants, with a tendency for their blood pressure to rise approximately 1 mmHg per year more quickly (Ref 12). Among male and female migrants, the age-adjusted prevalence of diabetes increased from 7.5 and 11.7 to 10.8 and 19.9 per 100, while it also climbed for non-migrants, from 3.0 and 8.7 to 7.0 and 14.3 per 100 (Ref 13). The researchers cited higher-protein diets, more alcohol, weight gain, altered physical activity, and food by fuel value as probable factors, not saturated fat (Ref 13). These represent correlations.

Two other populations with low recorded disease

In Tokelau, coconut provided 63 per cent of the dietary fuel, while it accounted for 34 per cent in Pukapuka. The higher saturated fat consumption, according to the authors, accounted for the 35 to 40 mg/dL increase in serum cholesterol levels among Tokelauans (Ref 14). This backs up the initial link described in the traditional sequence. Vascular disease occurred rarely in both groups, yet there was "no evidence of the high saturated fat intake having a harmful effect" (Ref 14). This raises doubts about the second connection, despite the fact that rare does not equate to nonexistent and no contemporary imaging was available.

The Tsimane in Bolivia offer imaging. Among 705 adults aged 40 or above, 85 per cent had a CT scan showing a coronary artery calcium (CAC) score of zero. Over 75 years old, 65 per cent maintained a CAC score of zero, while just 8 per cent reached 100 or higher, a figure the authors described as "a five-fold lower prevalence than industrialised populations" (Ref 7). The participants’ diet consisted of an estimated 14 per cent protein, 14 per cent fat, and 72 per cent carbohydrate; meat and fish were acquired via hunting and fishing, while men engaged in physical activity for 6 to 7 hours daily and women for 4 to 6 (Ref 7). The average LDL level reached 2.35 mmol/L, up from an average of 1.84 mmol/L during 2004-2011, a trend the authors attribute to the introduction of motorised access to market foods. The Tsimane, unlike the Kitavans, had a significant inflammatory load, as indicated by high-sensitivity CRP levels exceeding 3.0 in 51 per cent of individuals (Ref 7).

According to the Tsimane authors, their interpretation of the findings is distinct from mine. The authors determined that coronary atherosclerosis "can be avoided in most people by achieving a lifetime with very low LDL, low blood pressure, low glucose, normal body-mass index, no smoking, and plenty of physical activity", though they noted that the relative contributions "are still to be determined" (Ref 7). The factors that predicted calcium levels in their analysis included age, body fat, hs-CRP, and ESR, with the model accounting for just 19 per cent of the attributable risk (Ref 7). Such a population is incapable of determining the significance of individual factors, and CAC identifies solely calcified plaque.

The 2026 CT study, involving 1,232 Tsimane and Moseten individuals, revealed intracranial vascular calcifications in over 90 per cent of them, exceeding rates observed in similarly aged industrialised groups and linked to age, brain atrophy, impaired cognition, and parkinsonian symptoms (Ref 15). The authors proposed that "alternative factors may contribute to intracranial arteriosclerosis" (Ref 15). Just because coronary calcium levels are low doesn't imply that other arteries are unaffected.

The mainstream case for LDL

The perspective that LDL lacks causal significance represents a minority stance within the medical field. An expert group from the European Atherosclerosis Society examined meta-analyses encompassing over 200 cohort studies, Mendelian randomisation studies and randomised trials involving more than two million individuals, and found a "remarkably consistent dose-dependent log-linear association" between LDL exposure (assessed by the cholesterol it transports) and atherosclerotic cardiovascular disease (Ref 8). The panel determined that the evidence "unequivocally establishes that LDL causes ASCVD", noting that reducing LDL offers advantages only if the decrease in the cholesterol it transports matches the reduction in LDL particle count (Ref 8). A Mendelian randomisation meta-analysis involving 312,321 participants and nine genetic variants discovered that prolonged exposure to reduced LDL levels corresponded to a 54.5 per cent decrease in coronary heart disease risk for every mmol/L decrease (Ref 9).

The American Heart Association issued a recommendation stating that trials substituting saturated fat with polyunsaturated vegetable oil "reduced CVD by approximately 30%", while highlighting that replacing it with primarily refined carbs and sugars "did not reduce CVD in clinical trials", an observation with which I partially concur (Ref 16). The 2020 Cochrane review demonstrated that lowering saturated fat decreased combined cardiovascular events by 17 per cent, with a number needed to treat of 56 in primary prevention, yet showed minimal or no impact on all-cause or cardiovascular mortality, both classified as moderate-quality evidence (Ref 17). The European Atherosclerosis Society's second statement found "strong support for a causal association" between triglyceride-rich lipoproteins, their remnants, and the risk of myocardial infarction, ischaemic stroke, and aortic valve stenosis (Ref 18). Though it aligns with my metabolic perspective, this doesn’t simplify the Kitava situation, given that their triglyceride levels weren’t low (Ref 6).

I argue that a significant portion of this relies on observational and genetic findings. Mendelian randomisation relies on the assumption that variants influence disease exclusively via LDL, an assumption that cannot be fully validated, and a lifelong genetic difference does not equate to the exposure of a dietary change. The initial consensus statement employs the cholesterol transported by LDL as an indicator of particle count, acknowledging that the advantages rely on their concurrent changes. The positive AHA studies replaced saturated fat with industrial seed oils, which I regard as detrimental. Kitava and the Tsimane don’t fit neatly into a single continuous dose-response for LDL, but population comparisons come with their own confounding factors, so neither side should make excessive claims.

What this means in practice

In my interpretation, these groups indicate that a lipid profile alone is an inadequate measure, whether under- or overestimating. Most Tsimane displayed no coronary calcium despite elevated inflammation levels, while the Kitavans exhibited minimal cardiovascular disease even with an unfavourable triglyceride and HDL profile.

In my current understanding, the indicators that most strongly correlate with a lack of heart disease in these contexts are leanness, regular physical activity, low diastolic blood pressure, and, on Kitava, low fasting insulin levels. These findings indicate links rather than conclusive evidence. In my view, LDL has never been a concern for those with good metabolic health, and the key markers I focus on include fasting insulin, triglycerides, the triglyceride to HDL ratio, HbA1c, and inflammation. My clinical perspective is not validated by Kitava, and I am willing to consider evidence that contradicts it.

There is no long-term controlled study examining how a strict carnivore diet impacts hard cardiovascular outcomes, similar to the absence of such trials for most dietary approaches.

Disclosures

I operate a private clinic where I provide consultations for a fee, covering topics like metabolic health and cholesterol, which means I have a financial stake in readers benefiting from this analysis. My editorial stance, in contrast to the established guidelines and some referenced studies, maintains that LDL does not cause heart disease, and I consume and generally recommend an animal-based diet. I declare no financial stake in any food or product discussed.

This article is not tailored medical advice and should not replace the recommendations from your personal physician. Individuals taking lipid-lowering drugs must consult their prescriber before making any changes to their medication. If you're interested in assistance tailored to your situation, I provide consultations, although numerous other skilled clinicians are also available, and you're under no obligation to choose me.

References

  1. Lindeberg S, Lundh B. Apparent absence of stroke and ischaemic heart disease in a traditional Melanesian island: a clinical study in Kitava. J Intern Med. 1993;233(3):269-75. PMID: 8450295. https://doi.org/10.1111/j.1365-2796.1993.tb00986.x
  2. Lindeberg S, Berntorp E, Nilsson-Ehle P, Terent A, et al. Age relations of cardiovascular risk factors in a traditional Melanesian society: the Kitava Study. Am J Clin Nutr. 1997;66(4):845-52. PMID: 9322559. https://doi.org/10.1093/ajcn/66.4.845
  3. Lindeberg S, Nilsson-Ehle P, Terent A, et al. Cardiovascular risk factors in a Melanesian population apparently free from stroke and ischaemic heart disease: the Kitava study. J Intern Med. 1994;236(3):331-40. PMID: 8077891. https://doi.org/10.1111/j.1365-2796.1994.tb00804.x
  4. Lindeberg S, Eliasson M, Lindahl B, Ahren B. Low serum insulin in traditional Pacific Islanders: the Kitava Study. Metabolism. 1999;48(10):1216-9. PMID: 10535381. https://doi.org/10.1016/s0026-0495(99)90258-5
  5. Lindeberg S, Berntorp E, Carlsson R, et al. Haemostatic variables in Pacific Islanders apparently free from stroke and ischaemic heart disease: the Kitava Study. Thromb Haemost. 1997;77(1):94-8. PMID: 9031456. https://pubmed.ncbi.nlm.nih.gov/9031456/
  6. Lindeberg S, Ahren B, Nilsson A, Cordain L, et al. Determinants of serum triglycerides and high-density lipoprotein cholesterol in traditional Trobriand Islanders: the Kitava Study. Scand J Clin Lab Invest. 2003;63(3):175-80. PMID: 12817903. https://pubmed.ncbi.nlm.nih.gov/12817903/
  7. Kaplan H, Thompson RC, Trumble BC, Wann LS, et al. Coronary atherosclerosis in indigenous South American Tsimane: a cross-sectional cohort study. Lancet. 2017;389(10080):1730-1739. PMID: 28320601. https://doi.org/10.1016/S0140-6736(17)30752-3
  8. Ference BA, Ginsberg HN, Graham I, Ray KK, et al. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. Eur Heart J. 2017;38(32):2459-2472. PMID: 28444290. https://doi.org/10.1093/eurheartj/ehx144
  9. Ference BA, Yoo W, Alesh I, Mahajan N, et al. Effect of long-term exposure to lower low-density lipoprotein cholesterol beginning early in life on the risk of coronary heart disease: a Mendelian randomization analysis. J Am Coll Cardiol. 2012;60(25):2631-9. PMID: 23083789. https://doi.org/10.1016/j.jacc.2012.09.017
  10. Carrera-Bastos P, Fontes-Villalba M, Gurven M, Muskiet F, et al. C-reactive protein in traditional Melanesians on Kitava. BMC Cardiovasc Disord. 2020;20(1):524. PMID: 33334321. https://doi.org/10.1186/s12872-020-01812-7
  11. Agmon-Levin N, Bat-sheva PK, Barzilai O, Ram M, et al. Antitreponemal antibodies leading to autoantibody production and protection from atherosclerosis in Kitavans from Papua New Guinea. Ann N Y Acad Sci. 2009. PMID: 19758215. https://doi.org/10.1111/j.1749-6632.2009.04671.x
  12. Joseph JG, Prior IA, Salmond CE, Stanley D. Elevation of systolic and diastolic blood pressure associated with migration: the Tokelau Island Migrant Study. J Chronic Dis. 1983;36(7):507-16. PMID: 6874882. https://doi.org/10.1016/0021-9681(83)90128-5; and Salmond CE, Joseph JG, Prior IA, Stanley DG, Wessen AF. Longitudinal analysis of the relationship between blood pressure and migration: the Tokelau Island Migrant Study. Am J Epidemiol. 1985;122(2):291-301. PMID: 4014212. https://doi.org/10.1093/oxfordjournals.aje.a114101
  13. Ostbye T, Welby TJ, Prior IA, Salmond CE, Stokes YM. Type 2 (non-insulin-dependent) diabetes mellitus, migration and westernisation: the Tokelau Island Migrant Study. Diabetologia. 1989;32(8):585-90. PMID: 2777000. https://doi.org/10.1007/BF00285332
  14. Prior IA, Davidson F, Salmond CE, Czochanska Z. Cholesterol, coconuts, and diet on Polynesian atolls: a natural experiment: the Pukapuka and Tokelau island studies. Am J Clin Nutr. 1981;34(8):1552-61. PMID: 7270479. https://doi.org/10.1093/ajcn/34.8.1552
  15. Barisano G, Bigjahan B, Schleifer C, Ashna F, et al. High prevalence of cerebrovascular calcifications and clinical correlates in indigenous Bolivian forager-horticulturalists: a population-based observational study. J Am Heart Assoc. 2026;15(3):e043028. PMID: 41614317. https://doi.org/10.1161/JAHA.125.043028
  16. Sacks FM, Lichtenstein AH, Wu JHY, Appel LJ, et al. Dietary fats and cardiovascular disease: a presidential advisory from the American Heart Association. Circulation. 2017;136(3):e1-e23. PMID: 28620111. https://doi.org/10.1161/CIR.0000000000000510
  17. Hooper L, Martin N, Jimoh OF, Kirk C, et al. Reduction in saturated fat intake for cardiovascular disease. Cochrane Database Syst Rev. 2020;8(8):CD011737. PMID: 32827219. https://doi.org/10.1002/14651858.CD011737.pub3
  18. Ginsberg HN, Packard CJ, Chapman MJ, Boren J, et al. Triglyceride-rich lipoproteins and their remnants: metabolic insights, role in atherosclerotic cardiovascular disease, and emerging therapeutic strategies: a consensus statement from the European Atherosclerosis Society. Eur Heart J. 2021;42(47):4791-4806. PMID: 34472586. https://doi.org/10.1093/eurheartj/ehab551