Reply to: ‘Comment on “Microbiota Composition and Metabolism Are Associated With Gut Function in Parkinson’s Disease”’

We thank Dr. Zhang for his interest in our study1 and for the valuable feedback.2 Dr. Zhang correctly remarks that individuals in the same household typically share similar diets, highlighting our decision to enroll patient spouses as controls whenever possible. Control selection for case-control studies involves inherent trade-offs between comparability and efficiency,3 and we accordingly made substantial effort to control for potential confounders beyond simply selection strategy. This includes rigorous analyses of associations between diet and all variables of interest (see pp. 11–19 of the R-Markdown published as a supplementary file with our original article1). Dr. Zhang raises the interesting point that there may be relevant dietary variation in participants without a study-matched spouse, which is masked in the overall cohort when the 43 spousal pairs are included. Importantly, only 2 results presented in the article involved direct patient–control comparisons, where spousal–subgroup analysis might be relevant: (1) microbiota differential abundance (primarily performed to show validity with previous studies and largely presented as supplemental data) and (2) differences in microbial metabolite concentrations. To address Dr. Zhang’s point, we have repeated our dietary analyses in the subgroup of participants without study spouses (n = 154 patients with Parkinson’s disease and n = 60 controls). Confirming our observations in the full cohort, we found no significant differences between patients and controls in the consumption of any dietary items in this subgroup (all false discovery rate (FDR)-adjusted P > 0.78, see Methods section in our article1), and no significant correlations between dietary items and microbial genera (FDR-adjusted P > 0.41) or microbial metabolites (FDR-adjusted P > 0.86). Visualizing dietary intake by principal component analysis reveals no separation by group (Fig. 1A), with permutational multivariate analysis of variance (PERMANOVA) test confirming no significant difference in participant distribution by Parkinson’s disease × spouse group status (P = 0.53, 99,999 permutations). We have also repeated our primary analyses involving direct patient–control comparisons by spousal subgroup. Reassuringly, the results are highly consistent across groups (Fig. 1B,C), with diminished statistical power attributed to smaller sample size, especially in the spousal subgroup. Notably, as only 125/300 participants had metabolomics data and only 86/300 participants were spouses, the resulting overlap of complete study couples with metabolomics data was only n = 26 (ie, 13 pairs), and our study was not powered to detect statistically significant differences in groups this small (Fig. 1C). As noted previously, we made a substantial effort to demonstrate that diet was not confounding these relationships. Cohort studies by their nature involve sampling a subset of a population and inferring broader generalizability. Interestingly, many of the microbiota differences we observe in patients with Parkinson’s disease, including increased Akkermansia and Bifidobacterium and decreased Faecalibacterium and Lachnospiraceae, are repeatedly observed in other cohorts across multiple continents4, 5 despite significant geographical and dietary differences. We believe this supports the notion that consistent microbiota alterations—and by extension, the novel metabolomic and gastrointestinal function results reported in our study—are widely generalizable to the broader population with Parkinson’s disease, recognizing, as always, that further studies are needed. (1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. First Draft, B. Review and Critique. M.S.C.: 1B, 1C, 2A, 2B, 2C, 3A, 3B A.C.Y.: 1B, 1C, 3B E.G.: 1B, 1C, 3B K.S.: 1A, 1B, 1C D.K.: 1B, 1C L.H.F.: 1B, 1C M.M.: 1B, 1C N.R.: 1C, 3B T.H.: 2C, 3B B.B.F.: 1A, 1B, 1C, 2A, 2C, 3B S.A.C.: 1A, 1B, 1C, 2A, 2C, 3B

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.