Diet, Gut Microbiota, and Bone Health: Unlocking the Connections

A microbiota‑supportive dietary pattern was associated with lower fracture prevalence and modestly higher bone mineral density (BMD) in a large U.S. sample, identifying a potentially modifiable exposure relevant to osteoporosis risk.

In a cross‑sectional analysis of 9,385 U.S. adults from NHANES (2009–2020), dietary exposure was measured using two 24‑hour recalls and outcomes included odds of fracture, osteoporosis prevalence, and BMD at the femoral neck and lumbar spine. Multivariable models adjusted for age, sex, BMI, smoking, calcium and vitamin D intake, and relevant medications. The principal association showed roughly a 10% lower odds of fracture per 1‑unit higher DI‑GM score, with concordant, smaller reductions in osteoporosis prevalence.

The DI‑GM index operationalized habitual diet by scoring higher intake of fiber‑rich plant foods, fermented dairy and other fermented products, and unsaturated‑fat–rich items as beneficial, and assigning adverse scores to red and processed meats, refined grains, added sugars, and diets high in saturated fat. Two recalls were mapped to these components, summed into a continuous score, and categorized into quintiles for trend analyses.

Findings extended across multiple bone outcomes but were site‑specific. Higher DI‑GM was associated with lower odds of fracture (OR ≈ 0.90 per 1‑unit increase) and lower odds of osteoporosis (OR ≈ 0.94 per 1‑unit increase), and with higher femoral neck BMD (β ≈ 0.003 g/cm2 per 1‑unit increase). The femoral‑neck signal was most notable and appeared stronger in men; no clear association was observed for lumbar spine BMD. Effect sizes are modest at the individual level but consistent across adjusted models.

Biological plausibility is supported by likely mechanisms including increased short‑chain fatty acid production, modulation of systemic and local inflammation, and improved intestinal nutrient and mineral absorption. Key limitations temper causal inference: the cross‑sectional design, potential residual confounding, some self‑reported fracture and osteoporosis outcomes, and measurement error inherent to dietary recalls. These results demonstrate robust associations but do not establish causation.