When they compared these diet-quality scores with the incidence of hip fractures over the years, the study team found a significant difference in risk between women who scored highest on the AHEI scale and those who scored lowest.
There was a similar overall pattern among women based on their DASH and Alternative Mediterranean diet scores, but once researchers adjusted for other factors like body mass and physical activity, the difference was too small to rule out the possibility it was due to chance. Among women younger than 75, however, there were statistically meaningful differences in fracture risk tied to the highest versus lowest scores on all three diet indices.
Among the study’s limitations, the small number of hip fractures in men might have reduced the researchers’ ability to identify associations between fractures and diet. In addition, all the study participants in the analysis were white, which may make the results less generalizable to people of other ethnicities, the authors acknowledge.
The study was not a controlled experiment designed to prove that diet quality influences hip fracture risk directly. The study team adjusted for many other lifestyle and health factors that might play a role, and they note that people with higher diet-quality scores tended to have lower body mass and higher leisure time physical activity.
“Vitamins and minerals including calcium, vitamin D and vitamin K are essential for maintaining bone hardness and structure throughout the lifespan,” says Priya Khorana, an independent nutrition consultant in New York City who wasn’t involved in the study.
“Sure, bone integrity declines and composition weakens as people age, but this rate of decline can be attenuated with proper diets that include these key nutrients – among other foods,” she says in an email.
Reference
- Fung TT, Meyer HE, Willett WC, et al. Association between diet quality scores and risk of hip fracture in postmenopausal women and men aged 50 years and older. J Acad Nutr Diet. 2018 Feb 1. pii: S2212-2672(17)31850-6. doi: 10.1016/j.jand.2017.11.022. [Epub ahead of print]
