Sugar: Bad science ~ Great headlines

 Just three weeks ago, the headlines of a popular British newspaper carried the banner headline on its front page linking fruit juice consumption with elevated blood pressure. It was based on a study reported in the generally well-regarded scientific journal Appetite and was conducted by researchers in Australia[1]. The sample was small, just 146 subjects and their diets were analysed using a food frequency questionnaire. The authors created three groups according to their intake of fruit juice: “Rarely”, with fruit juice intake ranging from 0 to less than 3 times per month, “Occasionally” ranging from once per week to 5-6 times per week and “Daily” corresponding to once per day to more than 3 times per day. They examined blood pressure using two techniques that measure central blood pressure and brachial blood pressure. The authors found no statistically significant effect of fruit juice consumption on brachial blood pressure but they did find a statistically significant effect of fruit juice intake on central blood pressure (systolic only). The authors in their discussion note that this is the first study to examine the link between fruit juice consumption and hypertension and it should be noted that they do associate fruit juice intake with fructose intake. They go on to say: ”These findings are important as there is a common perception that fruit juice is healthy” and they continue, thus: “Thus, frequent fruit juice consumption may be contributing to excessive sugar intake, typical of the Western population, exacerbating the prevalence of hypertension.”

Now lets pause here for a minute, because this is quite a sensational claim. The authors have not examined total sugar intake in their study so they themselves cannot imply that higher fruit juice intake is associated with higher sugar intake. What about sugar from the sugar bowl, from jams and preserves, biscuits and cakes and of course soft drinks? In the absence of data on total sugar intake, the authors can conclude nothing whatsoever. The authors never mention this fact when they consider the limitations of the study. Rather, they regret that they were not able to statistically control for “total energy intake or total calories” (the same thing by the way). But they were able to control for weight height and other potentially confounding factors. Why could they not control for energy intake? To fail to report total energy intake and the fraction of that intake contributed by total sugar and added sugar is unfathomable in terms of scientific rigour. It renders this paper utterly useless and it casts a shadow on the editorial process of the journal where it was published. But it made great headlines and no doubt the Daily Mail story was carried around the world to warm the cockles of the beating hearts of the “sugar is toxic” school of thought.

Now contrast that paper with one published about the same time by a consortium led by David Jenkins a Canadian based world leader in the field of diet and health[2]. This group conducted a systematic review and meta analysis of studies, which have measured actual fructose intake and followed the subjects up for many years to monitor the development of hypertension. They identified 3,749 scientific papers with the words “fructose” and “hypertension” in the titles and immediately scrapped 3,723 because they were animal studies, cell studies, case studies, reviews rather that actual studies and so on. Of the 26 articles remaining, 25 were omitted because they had inadequate end-points (14), were a non-prospective (follow-up) cohort (7) or did not disclose total fructose intake (3). Theses omissions were all based upon strict rules that are internationally regarded as the “must follow” rules of systematic reviews. That left them with just three studies. However, these were very big studies, which involved 223,230 subjects who were perfectly healthy when recruited at baseline. They were followed up for an average of 11 years and 58,162 went on to develop clinical hypertension.  The subjects were categorized in to (fifths) quintiles of fructose intake. The authors found absolutely no evidence that the greater the quantity of fructose consumed from the lowest to the highest fifth was in any way associated with hypertension. The statistical models included all known confounding factors.

It should be noted that all three of the above papers were in the literature for the Australian group to go and find but apparently, they did not. The “sugar is toxic” die-hards will remember their study. But the scientific world will remain far more impressed by the Jenkins systematic review.

---

[1] Pase, MP et al (2014) Appetite 84, 68-72

[2] Jayalath VH et al (2014) J Amer Coll Nutrition, 33, 328-339