Shibata · Nutrition and health 1992 · narrative review of observational studies · n=516+

Nutrition for the Japanese elderly.

Cited 33 times in the scientific literature.

Level 4 - case-series / case-control

Synthesis of non-randomized observational cohorts and cross-sectional dietary comparisons

PubMed 1407826 · doi:10.1177/026010609200800312 · record verified 2026-08-29

What was done

The paper evaluated the relationship between nutritional intake and life expectancy in elderly Japanese populations based on three epidemiological datasets: a cross-sectional nutritional assessment of 94 centenarians (1972–1973); a 10-year prospective follow-up (1976–1986) of 422 urban residents aged 69–71 assessing diet and survivorship; and a comparative cross-sectional study using 24-hour dietary records between residents of Okinawa Prefecture (longest life expectancy) and Akita Prefecture (shorter life expectancy).

What was found

In centenarians (n = 94), the proportion of animal protein to total protein was higher than the contemporary average Japanese level. In the 422 urban older adults, high intakes of milk, fats, and oils showed favorable effects on 10-year survivorship, with survivors showing longitudinal increases in animal food consumption (eggs, milk, fish, meat). In the regional comparison, Okinawan residents consumed significantly higher levels of Ca, Fe, vitamins A, B1, B2, C, and higher proportions of energy from proteins and fats, with lower intakes of carbohydrates and NaCl, compared to Akita residents. No exact numerical estimates, confidence intervals, or p-values were reported in the abstract.

Why it matters

This work highlighted that adequate intake of animal protein, dairy, fat, and essential micronutrients—alongside lower sodium intake—is positively associated with longevity and survival in the Japanese elderly, challenging assumptions that strictly low-fat or low-protein diets drove longevity in this demographic.

Limits

The abstract provides no quantitative effect sizes, variance measures, or statistical test values. Findings are based on observational and ecological comparisons susceptible to healthy survivor bias, regional confounders, and unmeasured lifestyle factors. Dietary assessment via 24-hour records in the regional comparison is subject to measurement error and recall bias, and the sample size for that comparison is not specified.

Cited by