The global burden of diabetes is increasing, with the latest data from the International Diabetes Federation estimating a prevalence of 285 million in 2010, increasing to a staggering 439 million by 2030¹. The increased premature ill health, reduced life expectancy, and increased death rate related to the presence of diabetes, provide reason for concern, and make its prevention a major public health priority. That the primary prevention of diabetes is possible through lifestyle behaviours, including a healthy diet, has now been well established by intervention trials among people at high risk of diabetes². However, what the specific contribution might be of the intake of fruits and vegetables to diabetes prevention, is unresolved.

We hypothesised that the inconsistent findings^3-5 for an association between intake of fruit/vegetables and the risk of type 2 diabetes might be explained by the known imprecision and measurement error in assessing fruit/vegetable intake from selfreport methods such as the food frequency questionnaire. We thus aimed to examine the association of plasma vitamin C, an objective biomarker of fruit/vegetable intake, and the risk of future diabetes. We also examined the association of diabètes with self-reported fruit and vegetable intake.

The EPIC-Norfolk Study

The European Prospective Investigation of Cancer-Norfolk (EPICNorfolk) study is a population-based cohort study. Men and women aged 40-75 years, resident in Norfolk, U.K. were identified from family-doctor age-sex registers, and of 77,754 invited, 25,639 attended for a health check visit in 1993-97. Dietary intake was assessed by a semi-quantitative food frequency questionnaire and vitamin C was measured in blood (plasma) samples. After exclusions, 21,831 individuals (9,815 men; 12,016 women) were eligible for analysis. Participants were without known diabetes at baseline, and we ascertained new-onset diabetes (n=735) over 12 years of follow up⁶.

An association might be obscured by measurement error

There was a striking inverse association between plasma vitamin C concentration and the risk of new-onset diabetes, while the association with total fruit/vegetable intake was modest, and lacked the ‘dose-response’ effect seen with plasma vitamin C (Figure)6. The risk of diabetes in the top fifth (compared to bottom fifth) of the plasma vitamin C distribution was 62% lower (95% confidence limits 48% to 72% lower), while for a similar comparison for fruit/vegetable intake there was a modest 22% reduction in risk of diabetes (95% confidence limits 0% to 40% lower risk). These analyses accounted for other factors, such as potential healthier lifestyles among those who eat greater quantities of fruit/vegetable or have higher vitamin C levels [adjusting for age, sex, family history of diabetes, lifestyle factors (alcohol intake, smoking, physical activity), socio-economic status (education, social class), obesity (body mass index and waist circumference), and intake of vitamin supplements].

Conclusions

Evidence for the benefits of greater intake of fruits/vegetables for diabetes prevention is elicited strongly and convincingly when plasma vitamin C (a valid biomarker) is used as a proxy for fruit and vegetable intake. In contrast, the same association is attenuated and more modest when self-reported fruit/vegetable intake is used. Because fruit and vegetables are the main source of vitamin C, these findings suggest that eating even a small quantity of fruit/vegetables may be beneficial and that the protection against diabetes increases progressively with the quantity of fruit and vegetables consumed.