The role of dietary patterns in osteoporosis-related outcomes in older adults

Fractures are a major public health issue in older adults and have two major risk factors: low bone mineral density (BMD) and falls. Nutrition plays a critical role in the prevention of these outcomes. Most previous studies have investigated single nutrients, such as calcium and vitamin D but focussing on individual nutrients does not account for the complex interaction between nutrients consumed in the diet, and could result in an inability to detect a small effect from a single nutrient in this context. Consequently, the approach of using dietary patterns to account for the overall effect of diet has been recommended for bone research. Few cohort studies have investigated the associations of dietary patterns with fracture and BMD, and none have investigated the impact of dietary patterns identified using a posterior method on falls or falls risk. Therefore, this thesis aimed to identify the role of such dietary patterns in these osteoporosis-related outcomes in adults (‚Äöv¢‚Ä¢50 years). This thesis has two major components: a systematic review synthesising the literature reporting associations between empirically derived dietary patterns with BMD and fracture in healthy adults (‚Äöv¢‚Ä¢18 years); and a longitudinal study of Tasmanian Older Adult Cohort (TASOAC) identifying dietary patterns, their predictors and their associations with osteoporosis-related outcomes (falls risk, BMD and fracture) in older adults (‚Äöv¢‚Ä¢50 years). The systematic review and meta-analysis included 23 observational studies in the systematic review, 21 in the best evidence synthesis and 4 in a meta-analysis. Of these, 12 were cross-sectional, 10 were longitudinal and one was a case-control study. Key findings were: 1. A healthy pattern was associated with an up to 36% lower risk of hip fracture (for all 4 studies, risk ratio (RR) = 0.73 (95% confidence interval(CI) 0.56, 0.96); I\\(^2\\) = 95% and RR = 0.64 (95% CI 0.56, 0.73); I\\(^2\\)=67% in the subgroup of studies in which fracture was ascertained by medical records rather than self-report). 2. There was conflicting evidence for the associations of dietary patterns with BMD at any site, total body bone mineral content and total fractures, and for western diet and hip fracture. However, the evidence was consistent in that there were no detrimental associations between a healthy pattern and BMD at any site, total body bone mineral content and total fractures nor beneficial associations between a western pattern and of these bone outcomes. The second component was the longitudinal data from TASOAC (1098 older adults at baseline: mean age 63 years and 51% of women). TASOAC followed up participants at an average time of 2.6 years (n=875), 5 years (n=768), and 10.7years (n=567) from baseline. TASOAC data were used to determine the dietary patterns in this cohort at baseline, generate a score for each pattern and determine the associations of these pattern scores with participants' characteristics and osteoporosis-related outcomes. The key findings were: 1. Four dietary patterns identified were predominantly comprised of: fruit and vegetable pattern (vegetables, fruits, potatoes, breakfast cereals excluding muesli and porridge); animal protein pattern (red and processed meats, fish, poultry); snack pattern (snacks, sweets, nuts, condiments); western pattern(pizzas, hamburgers, meat pies, sweets). 2. Fruit and vegetable and snack pattern scores were lower but western and animal protein pattern scores higher in men and current smokers at baseline. The sex difference in animal protein score increased over time (p=0.012).Snack score was positively associated with age and physical activity at baseline (p<0.008 for all), but the effect of age lessened over time(p=0.035). Animal protein and western scores were negatively associated with age at baseline, but the effect on western scores reduced over time(p=0.001). Animal protein scores were lower in retired people. People living in socially disadvantaged areas had higher western scores. 3. Higher baseline fruit and vegetable pattern score was associated with lower falls risk z-score at baseline (˜í‚â§= -0.05 per standard deviation (SD) (95% CI-0.09, -0.01)). There were no associations of falls risk z-scores with the baseline scores of the other patterns. 4. Femoral neck and hip BMD reduced over time, but the decreases were less for each SD increase in baseline scores of animal protein and western patterns (p<0.02 for all). Lumbar spine increased over time and the increase was greater with higher baseline scores of fruit and vegetable, animal protein and western patterns (all ˜í‚⧠= 0.001 g/cm\\(^2\\)/year/SD of pattern score, all p<0.02). 5. Baseline scores of fruit and vegetable and snack patterns were associated with a higher risk of lumbar spine BMD increasing over ten years (RR=1.06and 1.05 respectively, p<0.05). 6. There were no associations between any dietary pattern and incident fractures. In summary, the data from the systematic review and the analyses in this thesis suggest that current dietary guidelines that recommend adhering to a healthy diet and avoiding a western diet are also appropriate for the optimising bone health and preventing osteoporotic fractures. The results also identified potential target groups for interventions to improve diet quality namely men, smokers, retirees, and those experiencing social disadvantage to inform the targeting of clinical and public health practice and future research. A fruit and vegetable dietary pattern may be beneficial for reducing a proxy measure of falls risk, the falls risk z-score. However, there is still a lack of longitudinal data examining the association of dietary patterns and incident falls. Increases in LS BMD can be a marker of degenerative changes in older adults, so associations of higher scores of fruit and vegetable, western and animal protein patterns with increases in lumbar spine BMD raises the possibility that dietary patterns could be implicated in the development of osteoarthritis. Future research is needed including: ‚Äö Confirmation of any relationship between dietary patterns and falls in prospective studies measuring incident falls. ‚Äö Studies to investigate the effect of nutrition on osteoarthritis of the spine, and potentially at other sites. ‚Äö Randomized control trials to confirm whether improving diet quality can improve bone density and reduce falls and fractures, though these are likely to be costly and logistically challenging.