In the largest study to date, we have profiled the effects of gut microbiome diversity and dietary fibre intake on longitudinal weight gain. We showed that long-term weight gain is only in part determined by an individual's genetic make-up and that low gut microbiome diversity is associated with a higher weight gain over time. Our results on longitudinal weight gain are consistent with several studies that have provided evidence of associations between the gut microbiome and cross-sectional measures of body weight.7, 12, 13

In this study, the lack of microbiome data at baseline precludes us from being able to assess if higher diversity is a cause or a consequence of higher weight gain. We note two possible interpretations for the data reported. On the one hand, the longitudinal human data presented here is that gut microbiome composition could contribute to weight gain independently of calorie intake, physical activity and other potential confounders (such as, use of proton pump inhibitors or antibiotics).28, 29 An alternative interpretation is that weight gain may be contributing to lower bacterial diversity. This second hypothesis requires that at a fixed level of caloric intake, the host metabolism leads to both higher weight gain and lower diversity. However, there is extensive evidence documenting that the microbiome composition influences energy metabolism30, 31 and at the same time, to our knowledge there are no proposed mechanisms for slower energy metabolism in the host influencing bacterial composition. If lower bacterial diversity was indeed directly linked to lower weight gain, this would be in agreement with what has been found in murine models regarding the effect of the gut microbiota on energy metabolism in the host31 and would suggest that gut microbes may be viewed as ‘novel’ future therapeutic target to treat obesity.

We report that microbiome diversity could be influencing the observed relation between dietary fibre and weight gain. When we stratified the sample between those in the top tertile of Shannon’s diversity and those in the bottom tertile, we found that fibre intake is significantly associated with a decreased risk of being in the high weight gain group among individuals in the high microbiome diversity group but not in those with low microbiome diversity.

We also identified nine OTUs to be significantly associated with weight gain. Adjusting for proton pump inhibitor and antibiotics did not change the results. Conflicting evidence exists regarding phylogenetic signatures in obese human guts, with many studies indicating and increased ratio of Firmicutes: Bacteoridates,13, 32, 33, 34 some showing no trend and some showing the opposite trend.35, 36, 37 Here we found that among the eight OTUs that are significantly associated with lower risk of weight gain, seven belong to the Firmicutes family, many of them part of the Ruminococcaceae. The suggestion that this ratio may not be particularly informative regarding the role of the microbiome in determining body weight had already been put forward by others.38

The association between some Ruminococcaceae and lower risk of weight gain and Bacteroides and higher risk of weight gain may be simply because of their (respectively) positive and negative correlations with microbiome diversity, although in two instances the OTUs remain associated even after adjustment for diversity.

In mice, the gut microbiota is altered during suppression of obesity in a cold environment. Ruminococcaceae Adlercreutzia and Desulfovibrio39 are among the bacteria that increase during this process. Thus, it is possible that Ruminococcaceae may be functionally linked to a lean phenotype but further functional studies are needed to assess if this is the case.

A small interventional study in 33 obese individuals identified significant microbiome changes, including a decrease in Faecalibacterium prausnitzii, under weight loss in 4 months.40 In our data, however, we find no significant association of Faecalibacterium prausnitzii with longitudinal weight change, although we find that Faecalibacterium prausnitzii correlates cross-sectionally with lower BMI (beta (s.e.)= −0.54 (0.11), P=1.4 × 10−6) consistent with an association between the abundance of this species in the gut and obesity. We note that we studied a normal population and not an obese group and that this study had a larger study sample and considerably longer follow-up time. However, this suggests that changes in the microbiome in response to weight loss over a short period of time (that is, 4 months) may not reflect differences in microbiome composition associated with lower risk of weight gain over a period of many years.

Not only is weight gain in large part because of non-genetic factors,1 but an individual’s gut microbiome diversity is only in part determined by the hosts' genetic make-up. The heritability of gut microbiome diversity has been estimated to range from 0.30 to 0.37,21 which means that over 60% of the variation in microbiome diversity is environmentally determined and understanding how to increase microbiome diversity should be a focus of future research.

Our results also suggest that the beneficial effect of fibre on weight may be more pronounced in individuals with higher microbiome diversity, although this may reflect at least in part the fact that individuals, which higher fibre have a greater microbiome diversity.41 The healthy effects of a diverse gut microbiome on several phenotypes have already been demonstrated in humans in various settings.42 Experimental work in animals has shown that fibre intake reduces the energy density of diet, and the resulting short chain fatty acids promote intestinal gluconeogenesis, incretin formation and subsequently satiety, whereas at the same time short chain fatty acids also deliver energy to the host and support liponeogenesis.43 Our data suggest that increasing microbiome diversity may be itself a desirable outcome and that an effect of fibre intake on reduced weight gain is seen more strongly in individuals with higher microbiome diversity.

We note several study limitations, the major one being the lack of measures of microbiome composition at baseline that would enable us to assess the predictive value of diversity with regards to weight gain. Smaller studies, however, have already shown that gut microbiome composition influence weight gain, for example, in children (n=25)44 and in individuals taking specific antibiotics (n=102)45 and hence our results are not only consistent but help better document, which OTUs are involved. Another limitation is that the population under study consists of women and there may be gender differences with regards to the role of the microbiome on weight gain. However, this is to our knowledge the largest study to date and the first to explore the association with weight gain over time and not just the association with obesity and leanness. Another limitation of our study is the type of dietary data available from FFQs, which being recall data are subject to some bias. For example, the lack of a significant association between protein intake and microbiome diversity may reflect the limits of FFQ recall data compared with those of carefully controlled dietary intervention studies, hence we cannot exclude the importance of protein intake either on weight gain or on the microbiome from these data.46

We also note that the measure of fibre used here referred only to total NSP as the more comprehensive measure was not available. According to the British Nutrition Foundation in the UK, the average intake of NSP is 12.8 g day–1 for women and 14.8 g day–1 with a recommended average intake for adults is 18 g (NSP) per day.47 In our data, the average dietary intake is 20 g NSP per day, which is therefore above the national average and in line with the British National Formulary recommendation.

In conclusion, this study is the first to correlate gut microbiome composition and diversity to long-term (intended as several years) weight change adjusting for calorie intake. It is also one of the largest studies to date linking obesity to the microbiome in humans. Our data are in agreement with other studies that support a role for the gut microbiome composition in the regulation of human body weight, which is to a large extent environmentally determined and independent of caloric intake. As the gut microbiome is modifiable, we believe these results should increase interest in targeting the microbiome for weight control interventions and should encourage research into longitudinal changes in the microbiome in sufficiently powered studies.