Osteoporosis, balance and nutrition in elderly women in primary care

Objective: Some people get skeletal fractures as a result of a trauma which wouldn’t cause a fracture in a healthy individual. These fractures are sometimes referred to as “fragility fractures”. Common sites for fragility fractures are the spine, hip, upper arm, shoulder, distal forearm and pelvis. Great effort has been put into the construction of different fracture risk assessment tools. The clinical importance of a correct estimate of fracture risk comes from the effective treatments which are available to lower fracture risk. The aim of study I was to investigate whether participants classified by MNA at risk of malnutrition or malnourished had a higher mortality risk than participants with normal MNA. The aims of study II were to investigate whether one-leg standing time (OLST) could predict hip fracture risk and to compare the predictive accuracy of OLST to that of the fracture-risk assessment tool FRAX. The aims of study III were to investigate whether there was an association between high serum concentration of IGFBP-I and high fracture risk, and to evaluate whether such an association could be mediated by either BMI or IGF-I. The aims of study IV were to investigate whether addition of either maximum gait speed or OLST could improve the predictive accuracy of FRAX, as measured by AUC, Harrell’s c and NRI.

Methods: Study population: All four studies were based on data from the same study population. Inclusion criteria were living in the Bagarmossen area south of Stockholm, Sweden, and being a woman born between 1920 and 1930. These criteria were met by 937 women in 1999. Two rounds of invitations were sent after which 351 women were included in the study. Bone mineral density measurements: Bone mineral density measurements were conducted in 1999–2001 using Hologic QDR 4500 DXA equipment (Hologic, Waltham, MD, USA). Calibration was performed daily with a phantom. Biochemical analyses: Serum IGFBP-I was analyzed with radioimmunoassay (RIA). Total IGF-I was analyzed in serum by RIA after separation of IGF-I from IGF-binding proteins by acid ethanol extraction and cryoprecipitation. Functional tests: OLST was measured twice on each leg with eyes open. The longest time out of the four was used for analysis. Gait speed was measured at a walk as fast as possible 15 meters down a corridor with a flat floor, turn at a mark and then walk back again to the starting point. The average speed was calculated. Vitamin D was analyzed with Nichols Advantage® 25-Hydroxyvitamin D assay (Nichols Institute Diagnostics), a chemoluminescence analysis.

Results: In study I the age-adjusted HR for mortality was 2.36 (1.25–4.46) for the “risk of malnutrition” group with MNA ≤23.5 compared to the “normal nutritional status” group with MNA >23.5. In study II the age-adjusted HR of a hip fracture was 0.95 (0.92–0.98) for one second longer OLST. Participants with an OLST <10 s had an age-adjusted HR for a hip fracture of 2.60 (1.36–4.95). In study III IGFBP-I had a positive linear relation to both the risk of a hip fracture and the risk of any major osteoporotic fracture. This relation was partly mediated by 45% by femoral neck BMD (Figure 3). However, the relation between IGFBP-I and hip fractures was not confounded by IGF-I, BMI or GFR as could be suspected. In study IV the age-adjusted HR for a hip fracture for gait speed <0.8 m/s compared to gait speed ≥0.8 m/s was 6.89 (2.87‒16.51). Categorical NRI for addition of gait speed to FRAX was 0.24 (p=0.023) when highest quartile of FRAX-estimated fracture risk was denominated as the high-risk group.

Conclusions: A MNA score ≤23.5 was associated to an increased mortality but the relation could be explained by the use of loop-diuretics which indicated heart- or kidney failure as the cause of death. Both OLST and gait speed were FRAX-independent predictors of both hip fractures and major osteoporotic fractures. Although OLST and gait speed were moderately correlated, none of them was a confounder to the other’s relation to fracture risk. IGFBP-I had a positive linear relation to the risk of both hip fractures and major osteoporotic fractures. IGFBP-I was neither correlated to nor confounded by OLST or gait speed. However, FRAX-estimated fracture risk seemed to be a confounder.

List of scientific papers

I. Lundin H, Saaf M, Strender LE, Mollasaraie HA, Salminen H. Mini nutritional assessment and 10-year mortality in free-living elderly women: a prospective cohort study with 10-year follow-up. Eur J Clin Nutr. 2012;66(9):1050-3.
https://doi.org/10.1038/ejcn.2012.100

II. Lundin H, Saaf M, Strender LE, Nyren S, Johansson SE, Salminen H. One-leg standing time and hip-fracture prediction. Osteoporos Int. 2014;25(4):1305-11.
https://doi.org/10.1007/s00198-013-2593-1

III. High serum insulin-like growth factor binding protein 1 (IGFBP-I) is associated with high fracture risk independent of insulin-like growth factor 1 (IGF-I). [Submitted]

IV. Gait speed and one-leg standing time each add to the predictive ability of FRAX. [Submitted]