In this population-based study of Finnish adults we examined the associations of plant-based foods, red and processed meat, and dairy and their subgroups with gut microbiome. We found that fruit consumption was positively and a subgroup of dairy including milk, cream and ice-cream was inversely associated with intra-individual diversity (alpha-diversity) and that plant-based foods (except for potatoes), dairy and their subgroups (except for fermented milk) were associated with compositional differences between individuals (beta-diversity) in microbiome. Plant-based foods, red meat and dairy were also associated with relative abundances of distinctive bacterial species. Furthermore, dairy and of the subgroups fermented milk, other dairy products, vegetables, and cereals were associated with microbial pathways.

Findings regarding alpha- and beta diversity from previous larger-scale observational studies are inconsistent. A French study on healthy adults (n = 862) found that fresh fruit consumption frequencies were positively and meat (not specified) or processed meat were inversely associated with alpha-diversity, whereas fresh/cooked vegetables, cooked fruits, legumes, dairy products, or cheese were not associated with alpha-diversity [11]. Of these fresh and cooked fruits and cheese, however, were associated with beta-diversity. A recent Chinese study on middle-aged and elderly people (n = 1879) found that fruit but not vegetable consumption (g/day) was associated with alpha and beta-diversity [14]. Another Chinese study (n = 702) also on middle aged and elderly found no association between consumption frequencies of fruits, vegetables, whole or refined grains, red meat or dairy products with alpha-diversity, whereas of these whole grains and vegetables were associated with beta-diversity [12]. In our previous study (n = 4930) fruits and berries consumption (frequencies) were positively associated with alpha diversity which is in line with our current study, however differing from this current study were the positive associations found also for fiber-rich breads (representing cereals) and low-fat cheese (representing dairy group) [13]. As for beta-diversity all the aforementioned foods, in addition to vegetables and red-meat products (unlike in the current study), were also associated with beta-diversity.

Geographical variations in microbial compositions across studies with diverse populations and food cultures may contribute to observed differences [38]. Another potential explanation could be related to the differences in utilized dietary assessment methods potentially influencing the findings to some extent.

In species level analyses, consumption of plant-based foods was positively associated with cellulolytic species such as Ruminococcus albus [39] and several known short-chain fatty acid (SCFA)-producing species from Lachnospiraceae family such as Eubacterium eligens, Roseburia hominis and Butyrivibrio crossotus [40]. Red and processed meat consumption was associated with only few individual bacterial species including a positive association with Clostridium disporicum, previously linked to secondary bile acid production [41] and an inverse association with R. hominis (accounted for by processed meat subgroup). Also our previous study [13] as well as the other aforementioned observational studies [11, 12, 14] have linked Ruminococcus, Roseburia or Eubacterium genera to higher consumption of several plant-based foods including fruits [11, 13, 14], vegetables [12, 13] or fiber-rich breads [13]. In the current study these associations were also seen in all subgroups of the plant-based foods (except for potatoes) in general. Furthermore, a US trial (n = 9) examining short-term effects of plant-based diet and animal-based diets on gut microbiome found that species from Ruminococcus, Roseburia and Eubacterium genera were more abundant with plant-based diet [7]. The SCFAs, mainly butyrate, acetate, and propionate, can affect human health including potential benefits for insulin sensitivity and management of type 2 diabetes and obesity [42]. Reduced fecal SCFA levels have also been associated with a higher risk of colorectal cancer [43]. Furthermore, we also found that the plant-based subgroups cereals and vegetables were associated with species from the Prevotella genus, which is in line with previous literature showing an association with plant rich diets low in animal-sourced foods [44], for example a US study on healthy participants (n = 98) showed a positive association of Prevotella genus with diets rich in plant-derived fibers and carbohydrates and inverse association with fatty and amino acid rich diets [45]. Cereals were also positively associated with several species from genus Bifidobacterium. This was also seen in a recent U.S. study on healthy adults (n = 343) [46] and in our previous study with fiber rich bread [13]. A systematic review on human intervention studies (n = 40 studies) for the effects of cereal fibers on gut microbiota composition in healthy adults indicated increases in Bifidobacterium species abundances by wheat, oat and barley but not with rye [47]. Furthermore, we found that dairy consumption was positively associated (mainly accounted for by fermented dairy subgroup) with several lactic acid bacteria including commonly used starter cultures for fermented dairy products such as Lactobacillus delbrueckii and Streptococcus thermophilus [48]. Inverse associations were found with several species of Prevotella genus. These findings are in line with a recent systematic review of randomized controlled trials (n = 468) on effects of dairy consumption on human gut microbiome which concluded that dairy consumption increased the abundance of the beneficial genera Lactobacillus and Bifidobacterium and there were also indications of reductions in Prevotella species [49]. Furthermore, we found that dairy consumption was positively associated with several species from Enterobacteriaceae family (mainly accounted for by other dairy products subgroup) such as Citrobacter freundii and Klebsiella oxytoca. Although these species have been recognized as natural species of microbiome related to dairy and dairy products, they may also be potentially opportunistic pathogens to humans [50,51,52].

Our pathway analysis indicated that of the main food groups dairy consumption may be associated with several microbial pathways including those related to carbohydrate metabolism and glycan biosynthesis and metabolism. Of the subgroups, similar pathway associations were observed with fermented milk and for the vegetables and cereals several associations were also detected in pathways related to metabolism including carbohydrate or lipid metabolism. Further studies are, however, needed to explore the importance of these findings and information is also needed on the possible causal mechanisms behind these associations.

The strength of our study included a population-based approach using a representative random population sample. We were also able to consider several crucial factors influencing the gut microbiome, however, there are certain factors we were unable to account for including the use of other medications (e.g., laxatives, proton pump inhibitors), recent gastrointestinal surgery or recent substantial diet alterations. Furthermore, whole metagenomic shallow shotgun sequencing utilizing SHOGUN optimized to be used with this kind of sequencing provides more robust taxonomic and functional information compared to 16 S RNA amplicon sequencing [20]. However, while superior to 16 S RNA amplicon sequencing, shallow shotgun sequencing is less accurate than deep sequencing in capturing genetic features, requiring some caution in interpreting functional results [53]. Furthermore, the cross-sectional design of this study cannot reveal causality. Also, dietary intake assessment relied on self-reported data, possibly affected by memory and reporting biases. The 48-hour dietary recalls covered two consecutive days, and there was likely some correlation between the days. However, dietary recalls provide more detailed insights into food types and quantities, compared to questionnaires and may be valuable in capturing temporal associations with the gut microbiome. This is of interest since, despite the perceived stability of the gut microbiome, rapid shifts can occur due to diet [7, 54]. The temporal gap between the stool collections and dietary recalls in our study was relatively brief, ranging from a matter of day(s) to at most a few weeks. As a limitation, dietary recalls may not be an accurate estimate of long-term dietary habits. It’s noteworthy, however, that the examined food groups are prevalent foods in Finnish diets, which increases the likelihood of detecting associations and strengthens the reliability of our findings with a potential to reflect even more enduring dietary patterns. Additionally, in FINDIET 2002, the 48-hour dietary recalls were compared with a 5-day dietary information (48-hour dietary recall and a 3-day food diary) within a subsample, which demonstrated comparable results [15]. A recent study on British adults similarly supported the effectiveness of a 48-hour recall in characterizing dietary patterns when compared to a 5-day food diary [55]. It is also known that health-conscious people are more likely to participate in health surveys, however, our participation rate was rather high (64%) and therefore the bias related to the healthy participant effect is likely to be small [56]. Furthermore, the geographical differences in gut microbiome composition may limit the generalizability of our results to some extent in different sociocultural, and environmental settings, for example [38].

In conclusion, we found that consumption of plant-based foods and dairy contributed to the compositional differences in gut microbiome in terms of beta-diversity and that plant-based foods, red and processed meat and dairy were associated with distinct overall microbial community compositions. Plant-based food consumption particularly was associated with a larger number of putative beneficial species on average. These findings indicate that a shift towards more sustainable and healthier diets with more plant-based foods and moderate dairy consumption could also benefit our gut microbiome.