Three studies by US researchers have found that eating at least three servings a week of whole fruits is associated with a 2% reduced risk of type 2 diabetes. Different types of fruit had different associated reductions in diabetes risk, according to NHS choices.
Top of the chart were blueberries (often touted as a superfood), and eating an additional three servings a week of these was associated with a 26% reduction in diabetes risk.
Eating more grapes and raisins; prunes; apples and pears; bananas and grapefruits was also associated with reduced diabetes risk, although the reductions in risk seen were smaller. Drinking three portions of fruit juice a week was associated with an increase in diabetes risk of around 8%.
The study has many strengths, including its large size and long follow-up period. But it relied on self-reported fruit intake, which can be inaccurate. In particular, this may be a problem for foods that are not eaten regularly.
The results of this study support the benefits of eating at least five portions of fruit and vegetables a day, with only one portion of juice a day counting towards your five portions.
Where did the story come from?
The study was carried out by researchers from the Harvard School of Public Health and other research centres in the US, UK and Singapore. It was funded by the US National Institutes of Health, and one author was supported by the National Heart, Lung, and Blood Institute.
The study was published in the peer-reviewed British Medical Journal, and is available to view for free (open access).
The results are reported fairly accurately in the main body of each news story, but the headline in the Express implies all fresh fruits reduce diabetes risk by a quarter. In fact, the figure only applied to blueberries (specifically a 26% reduction). When the results for all the fruits included in the study were pooled there was only a 2% reduction.
What kind of research was this?
This was an analysis of three prospective cohort studies that looked at whether there is a link between consuming different fruit and fruit juices and the risk of developing type 2 diabetes.
The researchers say that although a diet including more fruit is recommended to reduce risk of many chronic diseases, including type 2 diabetes, observational studies looking at the link have found mixed results.
One possible explanation is that different fruits with different fibre, nutrient and antioxidant contents and different glycaemic indices (how quickly a food can raise blood sugar levels) could have different effects.
Prospective cohort studies such as this are the most feasible way to study the long-term effects of diet on health outcomes. Randomised controlled trials of diet would provide more reliable results, but are not feasible as people are unlikely to agree to stick to a highly regulated diet for a long period (sometimes up to 25 years) just for the purposes of a study.
What did the research involve?
The researchers used data from three prospective cohort studies which had assessed fruit and juice intake in large populations of individuals (nurses or other health professionals) and followed them over time to see who developed diabetes. They looked at whether the type of fruit a person ate or their juice consumption was related to their risk of diabetes.
The three studies included were:
- the Nurses’ Health Study of 121,700 nurses, which started in 1976
- the Nurses’ Health Study II of 116,671 nurses, which started in 1989
- the Health Professionals Follow-up Study of 51,529 health professionals, which started in 1986
These studies assessed participants’ lifestyles, including food intake and health at the start of the study, and then every two years, by questionnaire. They managed to follow up about 90% of people in each study.
For the current analysis they excluded people who reported having diabetes of any kind (type 1 or 2 or gestational diabetes), cardiovascular disease or cancer at the start of the study. They also excluded those with incomplete information on fruit or juice intake, or whose reported energy intake appeared unfeasibly high or low, those who had no follow-up data, and those who were diagnosed with type 2 diabetes but the date at diagnosis was unclear. This left 187,382 people for analysis.
In all studies the participants were asked at the start of the studies about how often on average they ate specific foods, and how many standard portions they ate. Questionnaires were also sent to update dietary intake every four years.
Ten groups of fruits were assessed consistently from the start of the studies:
- grapes or raisins
- plums or apricots
- apples or pears
Watermelon was asked about sporadically during follow-up. Fruit juices assessed were apple, orange, grapefruit and other juices.
Questionnaire reported intake was compared with food diary records in a small sample of individuals from two of the studies. Some fruits showed an accurate relationship between the questionnaire and diary results (such as for bananas and grapefruit), but it was lower for some fruit (such as strawberries in men).
The participants were asked if they had been diagnosed with diabetes, and if they answered “yes” they were sent a follow-up questionnaire that asked about symptoms, diagnostic tests and diabetes drug use. They were considered to have diabetes if they reported:
- one or more classic symptoms of diabetes (excessive thirst, polyuria, weight loss and hunger) plus raised blood glucose levels,
- raised blood glucose levels on two different occasions in the absence of symptoms, or
- treatment with antidiabetic drugs.
The researchers then looked at whether fruit or juice intake was related to diabetes risk. Analyses were adjusted for factors that could influence the results (potential confounders), such as:
- body mass index
- multivitamin use
- physical activity
- menopausal status
- oral contraceptive use
- total energy intake
- general healthy eating
- family history of diabetes
Analyses of individual fruits were adjusted for the other fruits and fruit juice intake.
What were the basic results?
Overall, the participants were followed up for a total of 3,464,641 years, and 12,198 participants out of the 187,382 (6.5%) developed type 2 diabetes in this time. On average (median) the participants ate between zero and one serving a week of the individual fruits being assessed.
After taking into account potential confounders:
Every additional three servings per week of whole fruit consumption was associated with a 2% reduction in risk of developing diabetes (hazard ratio [HR] 0.98, 95% confidence interval [CI] 0.96 to 0.99).
Eating three servings per week of blueberries; grapes and raisins; apples and pears; bananas; and grapefruit was associated with a statistically significant reduction in risk of diabetes compared with eating less than one serving per week.
This reduction ranged from 5% for grapefruit or bananas, to 26% for blueberries (HRs ranging from 0.95 to 0.74).
Eating three servings per week of prunes; peaches; plums and apricots; or oranges was associated with a trend for reduction in risk, but this was not statistically significant.
Eating three servings per week of strawberries was associated with a slight non-significant increase in risk, while cantaloupe was associated with a 10% increase in risk (HR 1.10, 95% CI 1.02 to 1.18).
Drinking three servings per week of fruit juice was associated with an 8% increase in risk (HR 1.08, 95% CI 1.05 to 1.11).
How did the researchers interpret the results?
The researchers conclude that the association between fruit intake and type 2 diabetes risk varies depending on the fruit. They say that eating more of some whole fruits, particularly blueberries, grapes and apples, is significantly associated with a lower risk of type 2 diabetes, while drinking more fruit juice is associated with a higher risk.
This study found that eating more of some whole fruits was associated with a reduced risk of developing type 2 diabetes, but eating other fruits was not. It also found that drinking more fruit juice was associated with an increase in risk. The study has a range of strengths, including:
- its large size (almost 190,000 participants)
- long follow-up (more than 3,000,000 years in total across all participants)
- collecting data prospectively
- assessing diet at a number of time points, not just once
- taking into account a range of potential confounders
There are also some limitations, including that people had to report their own diet and diagnoses, and may not always have done so accurately. This may be particularly a problem for foods that are not eaten regularly. For example, the correlation between food diaries and the questionnaire for strawberry intake in men was not very good.
In their juice analyses the researchers did not look at the impact of the type of juice people drank, for example, whether this was freshly squeezed or from concentrate, or sweetened or not. Different types of juice could have differing effects.
Although the researchers tried to remove the effect of a large range of potential confounders they may still have an effect. This makes it hard to determine the exact effect of one small component of diet, such as an individual kind of fruit. On average (median) people ate between zero and one serving a week of the individual fruits being assessed. Although the study was very large, the numbers of people eating three servings of the individual fruits may have been relatively small.
It's also worth considering that the study included mainly health professionals of European descent, and results may not apply to other groups.
Overall, the results suggest that eating more of most whole fruit may help reduce diabetes risk. Current recommendations are that a healthy diet should include at least five portions of fruit and vegetables a day, with fruit juice only counting for at most one of these portions.