Interestingly enough, this study suggests that residual confounding is likely to account for much, if not all, of the observed associations between the beneficial associations it found between higher intakes of raw vegetables and CVD risk.
- When assessing the independent effect of raw and cooked vegetable intake, only raw vegetable intake showed inverse associations with CVD outcomes, whereas cooked vegetables showed no association.
- For models of CVD incidence and raw vegetable intake, covariate adjustment attenuated HR from 0.79 to 0.88, with the LR chi-squared statistic declining by 81.9%.
This substantial attenuation suggests that were the potential confounders measured perfectly, much, if not all, of the observed association with reported vegetable intake, would be explained by residual confounding.
That said, one cannot rule out the possibility of a true causal protective effect.
Raw and Cooked Vegetable Consumption and Risk of Cardiovascular Disease: A Study of 400,000 Adults in UK Biobank (open access)
frontiersin.org/articles/10.33…
#nutrition #diet
frontiersin.org/articles/10.33…
#nutrition #diet
• • •
Missing some Tweet in this thread? You can try to
force a refresh
