Long-term a posteriori dietary patterns and risk of hip fractures in a cohort of women

Eva Warensjö Lemming, Liisa Byberg, Håkan Melhus, Alicja Wolk, Karl Michaëlsson

Dietary pattern analysis is a useful tool to study the importance of food components in the context of a diet and how they relate to health and disease. The association between dietary patterns and fractures is at present uncertain. We aimed to study associations between dietary patterns and risk of hip fracture in the Swedish Mammography Cohort, including 56,736 women (median baseline age 52 years). Diet data was collected in food frequency questionnaires at two investigations and dietary patterns were defined by principal component analysis using 31 food groups. Information on hip fractures was collected from the Swedish National Patient Register. Multivariable adjusted hazard ratios (HR) with 95% confidence intervals (CI) were estimated in Cox proportional hazards regression analysis. The two patterns identified-the healthy and Western/convenience dietary patterns-were time-updated and analysed. During a median follow-up time of 25.5 years, 4997 women experienced a hip fracture. Hip fracture rate was 31% lower in the highest compared to the lowest quartile of the healthy dietary pattern [HR (95% CI) 0.69 (0.64; 0.75)]. In contrast, women in the highest compared to the lowest quartile of the Western/convenience dietary pattern had a 50% higher [HR (95% CI) 1.50 (1.38; 1.62)] hip fracture rate. Further, in each stratum of a Western/convenience dietary pattern a higher adherence to a healthy dietary pattern was associated with less hip fractures. The present results suggest that a varied healthy diet may be beneficial for the prevention of fragility fractures in women.