A large study in the BMJ suggests that eating whole fruit may reduce the risk of diabetes, and that some specific fruits may be more protective than others, but that increased intake of fruit juices may raise the risk of developing the disease.

This paper was based on data from three prospective cohort studies of American healthcare professionals. These studies followed the health outcomes of the study groups over many years, gathering detailed data about health and lifestyles. The Nurses’ Health Study (1984–2008) covering 66,105 women, the Nurses’ Health Study II (1991-2009), 85,104 women, and the Health Professionals Follow Up study (1986-2008), 36,173 men, together provided 3,464,641 person-years of follow-up.

The study population was predominantly of European ancestry, and participants were free of major chronic diseases at enrolment. Detailed questionnaires were completed every two years. These reported on all aspects of lifestyle. Diet assessment included reported consumption of fruit and fruit juices overall, but also specific questions on grapes or raisins, peaches, plums or apricots; prunes; bananas; cantaloupe melons; apples or pears; oranges; grapefruit; strawberries; and blueberries.

The fruit juices were also specifically listed as apple, orange, grapefruit or other juices. Questionnaires were validated against diet records where possible, but the authors acknowledge that recall may be a source of error. The study subjects were scored on their adherence to healthy lifestyles and given a healthy eating index score.

The incidence of self-reported diabetes was the study’s outcome measure. Self-reported diabetes was validated by a supplementary questionnaire and checked against physician records where possible;

97% of the checked diagnoses were confirmed as accurate. The diagnostic thresholds for diabetes were lowered in June 1998 and the studies took account of this. Overall 12,198 subjects developed type 2 diabetes.

The researchers analysed the data to allow for personal, lifestyle and dietary risk factors for diabetes.

After this adjustment the data showed a hazard ratio (HR) of 0.98

(95% CI:0.96-0.99) for those subjects who ate three servings a week of whole fruit. This equates to a small reduction of 2% in the risk of diabetes.

However, the same analysis of three servings a week of specific fruits showed a risk reduction of:

• 26% for blueberries HR 0.74,

(95% CI:0.66-0.83)

• 12% for grapes and raisins HR 0.88, (95% CI:0.83-0.93)

• 7% for apples and pears HR 0.93, (95% CI:0.90-0.96)

The different results do not relate to the glycaemic index/glycaemic load of the individual fruits. Taking three servings of fruit juice a day increased the chance of developing diabetes by 8%, HR 1.08, (95% CI: 1.05-1.11).

The authors acknowledged the limitations of measurement from self-reporting, the estimate of physical activity and the problems of grouping some fruits together. However, the findings were consistent across each of the cohorts studied.

The authors suggest that the juicing process may remove some beneficial fruit compounds, and also the fibrous material that slows sugar absorption. The sudden rises in sugar may be responsible for triggering changes that lead to type 2 diabetes.

Individual fruits are thought to have properties that might explain the relative protection. The protective effects of blueberries, apples, pears and grapes may be related to their anthocyanin and polyphenol content. These chemicals have been shown experimentally to enhance glucose uptake.

At present the advice in the UK is to have five portions of fruit or vegetables every day. There seems to be evidence from this paper that this good advice may be further refined to suggest including fresh apples, pears, grapes, and especially blueberries when possible for protection against developing type 2 diabetes.

However, patients might be advised to avoid regular intake of fruit juices as part of their  ‘five a day’ as this may increase their risk.

Dr Matthew Lockyer

• Muraki I, Imamura F, Manson JE et al. Fruit consumption and risk of type 2 diabetes: results from three prospective longitudinal cohort studies. BMJ 2013; 347:f5001