Although we cannot rule out residual or unmeasured confounding factors, the combined estimates from studies that controlled for season, BMI, or physical activity still showed significant or marginally significant inverse associations, suggesting that our results may not be entirely explained by bias due to confounding factors.
← all excerpts
Circulating levels of vitamin D, vitamin D receptor polymorphisms, and colorectal adenoma: a meta-analysis.
1
—
—