Blue Zones Reassessed: What Holds Up and What Does Not

What are the Blue Zones and where did the idea come from?

Blue Zones are geographic areas identified in demographic surveys as producing an unusually high share of people who reach 90 or 100. The label originated with the AKEA study, published by Michel Poulain, Gianni Pes and colleagues in Experimental Gerontology in 2004. That paper calculated an extreme longevity index, the percentage of people born in Sardinia between 1880 and 1900 who became centenarians, then applied Gaussian smoothing across the island's 377 municipalities. The cluster sat in the mountainous central-eastern interior.

Among centenarians born inside the zone, the ratio between the sexes was 1.35, against 2.43 across the rest of the island, an unusually even split at an age where one sex normally outnumbers the other several times over. The authors did not claim to know why. They wrote that the mechanism remained unknown, offered consanguinity and a narrowed genetic pool as one hypothesis, and noted that none of the world's longevity hot spots identified over the years had been fully validated.

Dan Buettner's November 2005 National Geographic cover story carried the idea to a general readership. By 2024 the demographers who identified the zones listed four: Okinawa in Japan, Ikaria in Greece, Ogliastra in Sardinia and Nicoya in Costa Rica. A set of shared habits, meaning plant-forward eating, constant low-intensity movement and social embedding, became the most repeated lifestyle prescription in consumer longevity.

What does the data-quality critique actually claim?

The critique holds that the regions reporting the most extreme ages are also the regions least able to document them. Saul Justin Newman, a researcher at the Australian National University's Biological Data Science Institute and later at the Centre for Longitudinal Studies at UCL, first posted the analysis to bioRxiv in July 2019, revising it most recently in November 2025. The paper remains a preprint, not published in a peer-reviewed journal.

The specific findings are sharp. In the United States, the state-by-state introduction of birth certificates was associated with a 69% to 82% fall in the number of supercentenarian records, meaning people recorded as reaching 110. Only 18% of exhaustively validated supercentenarians hold a birth certificate, and in the United States that figure is zero. Supercentenarian birthdates cluster on days divisible by five, the signature of age heaping in populations that estimate rather than record. In Italy, England and France, where vital registration is more uniform, remarkable longevity instead tracked poverty, shorter life expectancy, higher crime and fewer residents aged 90 and over. In England and France, old-age poverty rates alone accounted for more than half the regional variation in reaching a remarkable age.

Newman had already published the underlying mechanism in peer review. His 2018 paper in PLOS Biology showed that age-estimation and cohort-blending errors introduced at rates below 1 in 10,000 are sufficient to generate the apparent deceleration and plateau of mortality at extreme ages. Rare documentation errors dominate very small populations, and people over 105 are a very small population. In September 2024 the Annals of Improbable Research awarded Newman the first Ig Nobel Prize in Demography.

How much of the critique is independently corroborated?

The registration problem is documented independently of Newman's analysis. In 2010 a nationwide review of Japanese family registration records found 234,354 people listed as alive at 100 or older whose whereabouts could not be established, including roughly 77,000 who would have been at least 120 and 884 who would have been 150 or older. The audit followed the discovery that Sogen Kato, recorded as Tokyo's oldest man at 111, had been dead in his home for around three decades, and cases in which families continued drawing pensions on relatives long deceased.

That episode establishes one thing firmly and another not at all. It establishes that national registers can carry six-figure counts of impossible ages, and that a pension gives a household a financial reason to keep a dead relative on the books. It does not establish that any specific person in a named blue zone was misdated, and coverage of this dispute routinely merges the two claims.

What do the original blue-zone researchers say in response?

In October 2024 six researchers who identified the zones issued a joint statement rejecting the critique. The signatories were Gianni Pes of the University of Sassari, Luis Rosero-Bixby of the University of Costa Rica, Christina Chrysohoou and Christodoulos Stefanadis of the University of Athens, and Bradley and Craig Willcox of the University of Hawaii. Their objection is that regional statistics of the kind Newman used cannot speak to the individual-level age validation their own studies performed.

In Sardinia, they wrote, every centenarian's age in the zone villages was checked against civil status databases running back to 1866, handwritten ecclesiastical records available from the 17th century onward, and a genealogical reconstruction of village inhabitants, with siblings' dates cross-checked to rule out identity switches. One false supercentenarian was found that way. In Nicoya, ages came from the Costa Rican civil registry established in 1883, with immigrants and late-registered births excluded. The signatories also object on scale: the Sardinian zone is roughly 50,000 people across six mountain villages, while island-wide crime and poverty figures describe a population of 1.6 million, most of it urban.

Both positions can be partly correct. Validating the age of a named 102-year-old is a different exercise from explaining why a region's raw counts were high enough to attract attention in the first place. The critique lands hardest on the second question.

Is the longevity advantage still there?

The strongest evidence that blue-zone advantages are transient comes from blue-zone researchers. Michel Poulain and Anne Herm, writing in the Journal of Internal Medicine in 2024, found that Okinawa's population divides into two groups: generations born before the Second World War, who show a highly favourable mortality pattern, and those born after, whose mortality is now definitively higher than mainland Japan's. In 2002 Okinawa fell from 4th to 26th place among Japan's 47 prefectures in male life expectancy. That paper lists the plausible invalidation of the ages of some of the oldest residents among the candidate explanations, and calls for in-depth demographic age validation.

Nicoya shows the same shape. Rosero-Bixby, using a nationwide survival database of 550,000 Costa Rican adults alive at any point between 1990 and 2020, reported in Demographic Research in 2023 that Nicoyan men born in 1905 had 33% lower adult mortality than other Costa Ricans, while those born in 1945 had 10% higher. Surviving Nicoyans born before 1930 remain exceptionally long-lived. His conclusion was that extreme-longevity hotspots are probably transient and should be reassessed continuously.

Which lifestyle findings hold up on independent evidence?

Several habits associated with the zones carry evidence that does not depend on the zones at all, and each was established in populations with functioning vital registration.

• Mediterranean dietary pattern. PREDIMED assigned 7,447 Spanish adults aged 55 to 80 at high cardiovascular risk to a Mediterranean diet with extra-virgin olive oil, one with mixed nuts, or a reduced-fat control diet. Over a median 4.8 years the hazard ratio for myocardial infarction, stroke or cardiovascular death was 0.69 (95% CI 0.53 to 0.91) and 0.72 (95% CI 0.54 to 0.95) against control. The trial was retracted and republished in 2018 after randomisation irregularities at two of eleven sites; the figures above are the corrected estimates.
• Legume intake. The Food Habits in Later Life study followed 785 people aged 70 and over in Japan, Sweden, Greece and Australia for up to seven years and reported a 7% to 8% lower mortality hazard per additional 20 g of legumes per day (RR 0.92, 95% CI 0.85 to 0.99). No other food group predicted survival consistently.
• Routine physical activity. A 2023 dose-response meta-analysis in the British Journal of Sports Medicine pooled 196 articles covering 94 cohorts and more than 30 million participants, and found a relative risk for all-cause mortality of 0.69 (95% CI 0.65 to 0.73) at 8.75 mMET-hours per week, roughly the 150 minutes of moderate activity in current guidelines. The curve is steepest at the low end, the quantitative form of the observation that people in these regions walked everywhere rather than trained. The same non-linearity underlies our guidance on how much Zone 2 cardio is enough.
• Social connection. Holt-Lunstad and colleagues pooled 148 studies covering 308,849 participants in PLoS Medicine in 2010 and reported an odds ratio of 1.50 for survival among people with stronger social relationships (95% CI 1.42 to 1.59).

Only PREDIMED is randomised, and it tested a dietary pattern in high-risk Spanish adults rather than the menu of any island. The rest are observational and carry the same limits as any other predictive marker, including grip strength and regular sauna use.

How should a reader hold the Blue Zones idea now?

Treat the Blue Zones as a hypothesis-generating observation with a contested demographic foundation and a set of recommendations that stand or fall on other evidence. The claim that specific regions produce extraordinary numbers of centenarians is under serious challenge, and it is weakening on its own terms as the pre-war cohorts die.

The practical filter is to ask what a recommendation would rest on if the zones were removed from the argument entirely. Beans, walking and friends survive that test. Regional rituals and products sold on the authority of a village in Sardinia do not, and the scrutiny that applies to any supplement label applies here. None of the evidence above establishes that a diet or habit treats or prevents disease in an individual, and none should be used to change medical treatment without clinical advice.