A Scoping Review of Methods Used to Investigate Relationships between Dietary Intake and the Gastrointestinal Microbiome

Abstract The gastrointestinal (GI) microbiome is intrinsically linked to human health and disease, with dietary intake being a major modifiable variable contributing to microbial colonization and fermentation. Diet–microbiome studies are heterogeneous in the methods used to assess and record dietary intake, to sequence the microbiome data, and to analyze associations. To capture the extent of these inconsistencies, we performed a systematic scoping review to synthesize the information on methods used to assess and compare relationships between diet and the GI microbiome in human studies. Our search identified papers with metrics of both dietary intake and the GI microbiome (using either 16S rRNA or metagenomic shotgun sequencing) and specific diet–microbiome associative analysis. The databases searched were MEDLINE, EMBASE, CINAHL, Cochrane, and Google Scholar, and the search produced 22, 228 unique records after de-duplication. After full-text screening, 1,262 articles were selected for full extraction. A subset of 295 studies captured whole dietary intake and investigated associations with the GI microbiome. This subset assessed diet using 84 different food frequency questionnaires, 12 alternative surveys, and 4 types of diet recalls or diaries ranging from 1 to 14 days. Fifty-five percent (n = 163) of studies investigated habitual dietary intake, 35% (n = 101) investigated recent dietary intake, and 10% (n = 30) used methods to assess both. Eighty-one percent of studies employed 16S rRNA sequencing, with wide variation in extraction, sequencing, pre-processing, and annotation methods. Gaps in the reporting of the methods for each step of the sequencing and analysis process were evident, limiting future comparison of study outcomes. Microbiome-specific statistical methods were used in 11% of the subset—predominantly correlational methods that are not recommended for microbiome data. In conclusion, this review revealed extensive inconsistencies within the exploratory diet–microbiome studies, despite the existence of recommendations to assist researchers. Collaborative efforts to develop consensus in methods and reporting are needed to substantially progress the obtaining of high-quality evidence and meta-analysis in the diet–microbiome research field.