Knife in watermelon on wooden table

    Can eating too much fruit trigger gestational diabetes?

    13 April 2017

    Women who eat large amounts of fruit during pregnancy may have a higher risk of gestational diabetes.

    A new study found that women who ate lots of fruit during their second trimester were four times more likely to develop the disease.

    The raised risk was particularly associated with fruits high on the glycaemic index.

    However, those who ate more fruit also had a higher total intake of carbohydrates. This may be the real culprit (see our analysis below).

    The study, published in Nature, aimed to investigate the association between fruit consumption during the second trimester and the occurrence of gestational diabetes. It tracked the diets of 772 women for more than a year.

    Of the 772 participants, 169 were diagnosed with gestational diabetes during the study period.

    An increased likelihood of gestational diabetes was particularly associated with consumption of tropical and citrus fruits, along with fruits high on the glycaemic index, which include bananas, pineapple, apricots, kiwis, watermelon, dates and raisins.

    Gestational diabetes is a temporary condition during pregnancy but it raises the risk of type-2 diabetes in later life.

    Instant analysis
    This was a prospective cohort study evaluating the association of fruit consumption and the occurrence of gestational diabetes. This study design is insufficient on its own in establishing causality, but establishes correlation.

    Gestational diabetes is diabetes diagnosed in pregnancy usually between 24 to 28 weeks of gestation, when insulin resistance is at its highest, which was not evident before pregnancy and doesn’t persist longer than six weeks after childbirth.

    During the study period, 1,126 women were deemed eligible for inclusion, of which the data of 772 women was subsequently analysed.

    The patients were divided into four groups (quartiles one to four) which represented stratification in terms of fruit consumption. The kind of fruit consumed as well as its glycaemic index (GI) was noted. The glycaemic index, in simple terms, is a measure of how quickly a particular food can affect blood glucose levels. Foods higher on the list will affect it faster than those lower down on the list.

    Results suggest an increased risk of gestational diabetes with increased fruit consumption, with higher risk associated with the ingestion of higher glycemic index fruits. Citrus and tropical fruits were particularly associated with this increase. The data was analysed taking into account different variables.

    The results can be simplified (or oversimplified) as a fourfold increase in risk of diabetes in the group with the highest consumption of fruit. However, this is misleading, both in terms of the actual results and the way the data was analysed. This is in contrast to established advice that eating more fruit is better for health.

    A closer look at the results is more instructive. The groups that consumed more fruit also had a higher intake of carbohydrates (as well as the other macronutrients, protein and fat). Those consuming on average 259g and 389g of carbohydrates (quartiles three and four) had a two and three times increased risk of developing gestational diabetes, respectively, as compared to the other quartiles, where consumption of both was significantly lower. These same two groups were in addition consuming 425g and 710g of fruit on average (400g fruit equals about five a day) which is a not insignificant amount of carbohydrate.

    Considering the endocrine effects of pregnancy in addition to the metabolic pathways involved in carbohydrate metabolism, it is more than likely that it is the effect of a larger carbohydrate load in total as opposed to just higher levels of fruit that is the real issue.

    It is no accident that those with the highest consumption of total carbohydrate, which included the highest consumption of fruit (itself a source of carbohydrate), appeared at higher risk of diabetes. Pregnancy is a diabetogenic state (ie, a state that can predispose to the development of diabetes). There is a relative level of insulin resistance in pregnancy that reaches a maximum at 24-48 weeks of gestation, mediated by multiple placental hormones that insulin cannot counter, thus leading to the development of diabetes in susceptible patients, not all of whom will exhibit risk factors.

    The more insulin secretion is stimulated, the more the effects of insulin resistance will be noted in the form of destabilisation of blood sugar control.

    While fructose (the carbohydrate found in fruit) itself is implicated in insulin resistance and impaired function of pancreatic beta cells responsible for insulin secretion, most authorities do not consider the fructose in fruit itself to be an issue, simply as one cannot under ordinary circumstances ingest sufficient sucrose from fruit for it to be a major issue. This obviously does not apply to the ingestion of foods containing high-fructose corn syrup, which have been linked to a number of health problems.

    Like all studies, this paper has its weaknesses. For instance, food and fruit consumption was assessed by questionnaire, which introduces an element of bias, as patients could have underreported or exaggerated their food consumption, not to mention the fact that there wasn’t a deeper analysis of the relationship between total carbohydrate load and gestational diabetes as opposed to mere fruit consumption.

    It does, however, add to our knowledge on the impact of larger total carbohydrate loads on the risk of development of gestational diabetes.

    Take-home message: if there is a link between fruit consumption and gestational diabetes, it appears to be present for fruits with higher GI as well as citrus/tropical fruits. Berries appear not to increase the risk.

    More research, of course, is needed before any definitive recommendations can be made, but personally I would urge pregnant women to ensure that the bulk of their carbohydrates came from non-refined sources.
    Research score: 4/5