Frailty: A D-Ficiency Syndrome of Aging?

In this issue of JCEM, Ensrud et al. (1) report on the relationship between frailty and serum levels of 25-hydroxyvitamin D (25OHD) in a cross-sectional and longitudinal prospective cohort study of more than 6000 women aged 69 and older from the Study of Osteoporotic Fractures. The authors found that low levels of 25OHD in all individuals, i.e. at or below 15 ng/ml (37.5 nmol/liter) at baseline, were associated with the greatest risk of frailty (i.e. classification of intermediate or high frailty) as measured by a composite score. Even among nonfrail women at baseline, those with serum levels below 15 ng/ml were found to have a significantly greater risk of death or incident frailty. Not surprisingly, women with 25OHD concentrations at 15–19 ng/ml also were at slightly greater risk of these outcomes, compared with women between 20 and 29 ng/ml. More surprising was the finding that those women with levels at baseline above 30 ng/ml also had a greater prevalence of frailty. These data imply that there may be a U-shaped curve for serum 25OHD and frailty outcomes in older women such that both low and high levels (below 20 or above 30 ng/ml) may place older individuals at a greater risk of disability. The results of these findings, particularly with respect to levels below 20 ng/ml, are consistent with reports from other observational cohorts of elderly males and females and would imply that optimal serum 25OHD concentrations for preventing frailty range between 20 and 30 ng/ml (2–5). Although the Ensrud paper (1) implies that serum levels of 25OHD below 20 ng/ml (50 nmol/liter) are associated with a greater risk of frailty, longitudinal cohort studies such as this can be confounded by numerous variables and must be considered within the hierarchy of evidence levels. Indeed, correlation in epidemiological studies is not necessarily causation, and other observational studies have produced conflicting results regarding threshold levels of 25OHD and frailty, falls, and fractures in older individuals. Remarkably, there is a paucity of randomized controlled trials sufficiently powered to test the role of vitamin D in preventing frailty-related adverse outcomes. Nonetheless, a brief review of earlier trials and the observational data is necessary to fully appreciate the importance of the current report from Ensrud et al. (1) and in turn its significance in underscoring the need for federal funding of large randomized placebo-controlled trials testing the role of vitamin D supplementation (and increased serum 25OHD levels) in preventing frailty in our elderly populations. There are several caveats inherent in analyzing the relationship between 25OHD and frailty. First, it should be appreciated that a serum 25OHD level is simply an integrated measure of exposure to dietary vitamin D and sunlight that can vary over time. Despite numerous attempts to link ambient concentrations of 25OHD to chronic disorders such as malignancies, infectious diseases, autoimmunity, and fractures, observational studies have been inconsistent and inconclusive. Second, the term frailty (as defined in this paper by criteria related to objective measures such as grip strength, physical activity, and walking speed, as well as subjective determinants like self-perceived exhaustion) is a global index of overall health status. Logically, it follows that frailty, as a composite measure, represents the inevitable consequence of aging, such that the very old as well as the most chronically ill are certain to score poorly on indices of physical performance. Similarly, infirm individuals often do not go outside; thus, solar exposure is minimal and many have poor nutrition, thereby leading to lower levels of 25OHD. Both confounding and reverse causation bias (illness or infirmity leading to low 25OHD levels, rather than low levels causing illness) are major concerns in observational studies and could lead to spurious associations. Proving cause (i.e. low 25OHD) and effect (i.e. a patient-specific outcome such as frailty or fracture) is bolstered by biological plausibility and the availability of randomized controlled trials, the highest level of evidence for inferring causality. In regard to the former, very low vitamin D states such as rickets or osteomalacia are commonly associated with muscle weakness and bone pain, suggesting that vitamin D is essential for the maintenance of skeletal muscle function as well as for bone mineralization. However, there remains controversy as to the level of expression of the vitamin D receptor in mature muscle cells and whether there is a direct effect of circulating vitamin D on muscle function, particularly in older adults (6). With respect to the highest levels of evidence, one recent randomized placebo-controlled trial by Sanders et al. (7) not only failed to demonstrate a positive effect of high-dose vitamin D administered yearly for 3 yr on either falls or fractures but surprisingly revealed that supplemented individuals had a greater risk of falls with serum levels of 25OHD above 30 ng/ml. Another recent randomized trial of vitamin D and calcium in more than 3000 Finnish women ages 65–71 yr also failed to show a positive effect of 800 IU vitamin D supplementation on fractures, despite serum 25OHD levels that rose from 50 to 75 nmol/liter (30 ng/ml) (8). In sum, considering the totality of evidence from observational studies, clinical trials, and the meta-analyses of vitamin D trials, the results are still conflicted with respect to frailty and 25OHD and depend on the number and type of trials included (or excluded), age of participants, duration of treatment and regimen used, compliance, and the effects of dietary supplements which have to be considered when measuring total vitamin D exposure. Is there a threshold effect of 25OHD and, if so, is there a safe maintenance level that should be established for the general population? In the 2009 Agency for Healthcare Research and Quality report pro