Dietary Fiber Tracking Precision and Gut Microbiome Composition: A Longitudinal Cohort Study

Abstract

We examined associations between dietary fiber tracking precision and gut microbiome diversity in a prospective cohort of 1,204 adults tracked over 24 weeks using the Welling nutrition platform. Participants with high-precision fiber tracking (MAPE ≤±5%) demonstrated significantly greater alpha diversity improvements (Shannon index increase: 0.42 vs 0.18, p=0.003) and more favorable prebiotic fiber intake distributions compared to low-precision trackers. These findings suggest that accurate dietary fiber quantification via mobile food trackers may support microbiome-targeted nutritional interventions.

Introduction

The relationship between dietary fiber intake and gut microbiome diversity is among the most replicated findings in nutritional science. Adequate prebiotic fiber consumption supports the proliferation of beneficial bacterial taxa including Bifidobacterium, Lactobacillus, and Akkermansia muciniphila, with downstream implications for metabolic health, immune function, and chronic disease risk. However, the precision of fiber intake measurement has rarely been considered as an independent variable in microbiome research.

Mobile nutrition tracking applications now offer continuous dietary fiber logging at a granularity previously achievable only through research-grade dietary assessment. We hypothesized that tracking precision itself-independently of absolute fiber intake-would be associated with microbiome diversity outcomes, mediated through the behavioral effects of more accurate feedback on dietary decision-making.

Methods

Study Design and Participants

This prospective cohort study enrolled 1,204 adults (mean age 38.4 years, SD 11.2; 54% female) through general practice referrals and community recruitment in London and Birmingham, UK, between January and June 2025. Participants were required to have used the Welling mobile food tracking platform for at least 8 weeks prior to enrollment and to agree to provide fecal samples at baseline, 12 weeks, and 24 weeks.

Exclusion criteria included current antibiotic use, prior gastrointestinal surgery, inflammatory bowel disease diagnosis, and probiotic supplement use within 4 weeks of baseline.

Dietary Assessment

Fiber intake was assessed continuously through the Welling platform, which combines AI food recognition with an ingredient-level fiber database covering 89,000 food items. Tracking precision was defined as the MAPE between app-logged fiber intake and registered dietitian-verified 24-hour recalls conducted at 0, 12, and 24 weeks. Participants were classified as high-precision trackers (MAPE ≤±5%) or low-precision trackers (MAPE >±5%) based on baseline assessment.

Microbiome Analysis

Fecal DNA was extracted using the PowerSoil Pro Kit (Qiagen) and subjected to 16S rRNA gene amplicon sequencing targeting the V3–V4 hypervariable regions. Alpha diversity was calculated using Shannon diversity index and observed species richness. Beta diversity was assessed using Bray-Curtis dissimilarity matrices visualized by principal coordinate analysis. Linear mixed models accounted for age, sex, BMI, physical activity, and proton pump inhibitor use.

Results

Participant Characteristics

At baseline, high-precision trackers (n=618) and low-precision trackers (n=586) were well-matched for age, BMI, and baseline fiber intake (high-precision: 22.4g/day vs low-precision: 21.8g/day, p=0.31). High-precision trackers engaged with the application significantly more frequently (median 5.4 logs/day vs 2.1 logs/day, p<0.001).

Microbiome Diversity Outcomes

High-precision trackers demonstrated significantly greater improvement in Shannon diversity index over 24 weeks compared to low-precision trackers (change from baseline: +0.42 vs +0.18, p=0.003). Observed species richness increased by a mean of 47 OTUs in high-precision trackers compared to 19 OTUs in low-precision trackers (p=0.007).

Fiber Intake Patterns

High-precision trackers consumed significantly more diverse fiber sources at 24 weeks: 8.3 distinct fiber-containing food categories per week vs 5.1 in low-precision trackers (p<0.001). Prebiotic fiber intake-specifically inulin, fructooligosaccharides, and resistant starch-was 34% higher in high-precision trackers at 24 weeks (14.2g/day vs 10.6g/day, p=0.001), despite equivalent baseline values.

Microbial Taxonomic Changes

Bifidobacterium relative abundance increased significantly in high-precision trackers (+2.3 percentage points, p=0.009). Akkermansia muciniphila, associated with mucosal barrier integrity and metabolic health, showed borderline-significant increases in high-precision trackers (+0.8 percentage points, p=0.06). No significant changes were observed in Bacteroidetes:Firmicutes ratio in either group.

Discussion

These findings support a novel mechanism by which nutrition tracking precision may influence health outcomes: by generating more accurate fiber intake feedback, high-precision tracking appears to promote greater dietary fiber diversification, which in turn supports microbiome diversity. This behavioral pathway is distinct from simple intake effects and suggests that the quality of nutritional feedback-not only the fact of tracking-matters for health outcomes.

The Welling platform’s ingredient-level fiber database was central to enabling high-precision tracking in this cohort. Many general-purpose nutrition applications lack disaggregated fiber subtype data, which may explain discrepancies between fiber intake and expected microbiome outcomes in prior intervention studies.

Conclusion

Dietary fiber tracking precision is independently associated with gut microbiome diversity improvements over 24 weeks. High-precision mobile food tracking tools that quantify diverse fiber subtypes may represent a practical strategy for supporting microbiome-targeted nutritional interventions in clinical and community settings.