Abstract

Background

Although both vitamin D (25-hydroxyvitamin D [25(OH)D]) insufficiency and the frailty syndrome are more prevalent in women than men, sex-specific associations have not been explored. We estimated sex-specific associations of low 25(OH)D with frailty. Vitamin D insufficiency can result in hyperparathyroidism, and thus, parathyroid hormone (PTH) was explored as a potential mediator in the relationship between 25(OH)D levels and frailty.

Methods

The sample included 444 male and 561 female participants aged 65 years and older from the InCHIANTI study for whom 25(OH)D levels and frailty information were available. Frailty was defined as the presence of at least three of the five following criteria: slowness, weakness, low energy expenditure, exhaustion, and weight loss. Logistic regression models estimated the association between serum levels of 25(OH)D and PTH with frailty, controlling for potential confounders.

Results

Independent of covariates, men with 25(OH)D <50 nmol/L had greater odds of frailty than those with 25(OH)D ≥50 nmol/L (odds ratio [OR] = 4.94, 95% confidence interval [CI] = 1.80–13.61). In women, the adjusted OR for frailty (95% CI) was 1.43 (0.58–3.56). The 25(OH)D ORs differed between men and women (p = .041). ORs changed little after controlling for PTH. However, when low energy expenditure was excluded from the frailty definition, adjusted OR for frailty in men (95% CI) was 2.18 (0.59–8.04); controlling for PTH attenuated this OR by 32%. In women, the OR (95% CI) for frailty (low energy expenditure excluded) was 1.54 (0.31–7.58) and was attenuated by 6% after controlling for PTH.

Conclusions

Vitamin D insufficiency was associated with frailty in men, but not in women. Results suggest that PTH mediates the relationship between 25(OH)D and nonenergy expenditure aspects of frailty.

FRAILTY has been described as “a biologic syndrome of decreased reserve and resistance to stressors, resulting from cumulative declines across multiple physiologic systems and causing vulnerability to adverse outcomes” (1). The syndrome is associated with incident falls, functional limitation, disability, and mortality (1,2); thus, preventing frailty may help slow progression of the disablement process in older persons.

Low 25-hydroxyvitamin D [25(OH)D] levels are common in older adults (3,4) and have been linked to falls (5,6), fractures (5), pain (3,5,7), sarcopenia (4), poor physical function (8,9), disability (3), and frailty (10). Further, vitamin D supplementation has been shown to improve physical function (11,12).

The importance of vitamin D in maintaining calcium homeostasis is well known. Cholecalciferol is synthesized in the skin in response to ultraviolet light or is ingested through food (eg, fatty fish, eggs, fortified dairy products) and hydroxylated into 25(OH)D in the liver. Older persons often have low 25(OH)D levels owing to age-related decreased efficiency of hydroxylation and reduced sunlight exposure (3,13). When 25(OH)D levels are low, active metabolite 1,25-dihydroxyvitamin D [1,25-(OH)2D] and calcium absorption decrease. The reduced serum calcium causes parathyroid hormone (PTH) levels to rise to stimulate 1,25-(OH)2D production, resulting in increased bone turnover and hip fracture risk (3,13). Emerging research suggests that vitamin D affects muscle strength and function (4,8), both directly and indirectly through PTH regulation and inflammation (3,4,9,14,15).

Many sequelae of low 25(OH)D (3,4,8–10) are considered dimensions of the frailty syndrome (1,2); thus, low 25(OH)D may be a reversible cause of frailty. Also, low vitamin D may affect men and women differently. Age-related levels of 25(OH)D decline earlier and faster in women than in men (16), which may explain the greater prevalence of frailty in women versus men. Whether the pathway to frailty differs by sex is unknown.

In this study, we examined sex-specific associations between low 25(OH)D and prevalent frailty in community-dwelling men and women aged 65 years and older. We also explored PTH as a potential mediator in the pathway between low 25(OH)D and frailty.

METHODS

Study Population

InCHIANTI is an epidemiological study with participants enrolled from residents of two Italian towns (Bagno a Ripoli in Tuscany, Greve in Chianti). The Italian National Research Council of Aging Ethical Committee approved the study. All participants provided informed consent. Details regarding design and sampling are available elsewhere (17). Participants responded to a home interview, participated in a medical examination, provided blood samples, and underwent functional performance tests.

The study randomly sampled 1,260 men and women aged 65 years and older (and additionally oversampled those aged ≥90 years), of whom 1,154 (91.6%) participated. Analyses presented here utilized data from 1,005 individuals who provided blood samples and information on at least one frailty criterion. Among included participants, 980 provided complete frailty information, whereas the remaining 25 provided partial information.

Frailty

Frailty phenotype was defined according to five dimensions proposed by Fried and colleagues (1): “shrinking,” “exhaustion,” “low energy expenditure,” “slowness,” and “weakness.” Each dimension was operationalized using previously published methods (18). “Shrinking” was measured as self-reported unintentional weight loss more than 4.5 kg (10 lbs) within the past year. Presence of exhaustion was indicated by a response of “occasionally” or “often/always” to the statement, “I felt that everything was an effort.” This statement is an item from the Center for Epidemiological Studies—Depression scale, which was validated in Italian (19). Participants reported their level of daily leisure physical activity in the past year; “low energy expenditure” was defined as either complete inactivity or performing low-intensity activities less than 1 h/wk. “Slowness” was defined as usual walking speed in the slowest quintile within groups defined by sex and height. Walking speed was measured on a 4-m course using photocell recordings at the course start and finish. The final measure averaged two walks. Grip strength was measured using a handheld dynamometer (Nicholas Muscle Tester; Sammon Preston, Inc., Chicago, Ill) by a standard method; “weakness” was defined as grip strength in the lowest quintile within groups defined by sex and body mass index (BMI). “Frailty,” “intermediate frailty,” and “robustness” were defined, respectively, as presence of at least three, one or two, and zero of the five criteria.

Biological Markers

Fasting blood samples were drawn and were processed and stored at −80°C until analysis. Serum 25(OH)D was measured by radioimmunoassay (RIA kit; DiaSorin, Stillwater, Minn). Intra- and interassay coefficients of variation (CVs) were 8.1% and 10.2%, respectively. Serum intact PTH was measured with a two-site immunoradiometric assay kit (N-tact PTHSP; DiaSorin). Intra- and interassay CVs were less than 3.0% and 5.5%, respectively. The samples were not measured in duplicate. Impaired renal function can affect 25(OH)D levels; therefore, creatinine clearance, calculated using the Cockcroft-Gault formula, was included in statistical analyses. Serum creatinine for this calculation was measured using a standard Jaffe method (Roche Diagnostics, GmbH, Mannheim, Germany).

Other Covariates

The following covariates were included in statistical analyses: age (years), education (years of schooling), blood collection season, smoking (pack-years), alcohol consumption, cognitive function, comorbidities, and BMI. Alcohol consumption was measured as number of drinks per week using the European Prospective Investigation into Cancer and Nutrition (EPIC) questionnaire (20). The Mini-Mental State Examination (MMSE) (21) was used to measure cognitive function. Presence of comorbidities was based on adjudicated measures combining information from self-report, medical records, and clinical examination. Comorbidities considered were congestive heart failure, peripheral arterial disease, hypertension, diabetes mellitus, osteoarthritis, myocardial infarction, angina, and chronic obstructive pulmonary disease. These conditions were previously found to be associated with low 25(OH)D and/or frailty and, hence, may be confounders (10).

Statistical Analysis

Vitamin D insufficiency for analyses was defined as 25(OH)D less than 50 nmol/L, consistent with published reports (3,14). Additional analyses included 25(OH)D as a continuous variable because no consensus on a cutoff exists, and recent research suggests that 25(OH)D levels needed to prevent secondary hyperparathyroidism may depend on age (16). High PTH was defined as the upper quartile of PTH concentrations (>32.4 ng/L).

Descriptive analysis included χ2 or Fisher exact tests for categorical variables and t tests or Wilcoxon rank-sum tests for continuous variables. Multinomial logistic regression was used to estimate covariate-adjusted odds ratios (ORs) of frailty category by 25(OH)D status and assess the interaction between sex and 25(OH)D status. Two sex-stratified models were estimated. In addition to covariates, Model 1 included 25(OH)D and Model 2 included 25(OH)D and PTH. Additionally, logistic regression was used to estimate adjusted ORs of each frailty criterion by 25(OH)D and PTH status.

RESULTS

The study included 561 women (55.8%) and 444 men (44.2%) with average (standard deviation) age 75.6 (7.6) and 74.2 (7.0) years, respectively. Prevalence of frailty was 11.8% in women and 8.9% in men; 43.6% of women and 35.6% of men were intermediate frail. The median (interquartile range) of 25(OH)D for women and men were 33.4 (22.7–49.4) nmol/L and 48.5 (34.1–73.1) nmol/L, respectively.

Table 1 shows study sample characteristics. Frail participants were older; had fewer years of education, lower MMSE scores, lower creatinine clearance and 25(OH)D, higher serum PTH; and were more likely to have congestive heart failure and chronic obstructive pulmonary disease compared with those of the same sex who were not frail. Frail women consumed less alcohol and were more likely to have peripheral arterial disease, hypertension, and angina than robust women. The most common frailty criterion for frail men and women was slowness.

Table 1.

Characteristics of InCHIANTI Participants Included in This Study

Characteristic Men
 
Women
 
Robust (N = 242) Intermediate Frail (N = 151) Frail (N = 39) p Robust (N = 243) Intermediate Frail (N = 237) Frail (N = 64) p 
Age (yrs), mean (SD71.7 (5.5) 76.2 (7.1) 80.8 (7.8) <.001 72.9 (6.1) 76.1 (7.5) 82.3 (7.4) <.001 
Education (yrs), mean (SD6.5 (3.6) 5.9 (3.7) 4.6 (2.7) .003 5.1 (2.7) 4.5 (2.8) 4.2 (2.8) .025 
Season of blood collection, number (%)         
    Spring (March–May) 40 (16.5) 30 (19.4) 8 (20.5)  63 (25.9) 34 (14.4) 18 (28.1)  
    Summer (June–August) 27 (11.2) 20 (12.9) 9 (23.1)  18 (7.4) 27 (11.4) 9 (1.6)  
    Fall (September–November) 70 (28.9) 50 (32.3) 13 (33.3)  71 (35.7) 102 (43.0) 26 (40.6)  
    Winter (December–February) 105 (43.4) 55 (35.5) 9 (23.1) .17 91 (37.4) 74 (31.2) 11 (17.2) <.001 
Smoking (pack-years), mean (SD22.7 (23.5) 25.8 (24.2) 29.0 (31.2) .22 2.0 (6.4) 3.3 (9.0) 4.4 (15.4) .095 
Alcohol consumption (drinks/wk), mean (SD14.0 (13.4) 12.4 (14.2) 8.8 (9.0) .075 4.2 (5.7) 3.0 (4.7) 3.3 (5.5) .030 
BMI (kg/m2), mean (SD27.0 (3.4) 27.2 (3.3) 26.7 (2.8) .61 27.5 (4.3) 27.8 (4.5) 28.3 (6.1) .42 
MMSE score, mean (SD26.4 (2.7) 24.2 (5.3) 20.9 (7.6) <.001 25.2 (3.6) 23.8 (5.0) 21.6 (5.4) <.001 
Congestive heart failure, number (%) 8 (3.3) 11 (7.1) 7 (18.0) .003 1 (0.4) 17 (7.2) 11 (17.2) <.001 
Peripheral arterial disease, number (%) 15 (6.2) 15 (9.7) 6 (15.4) .11 3 (1.2) 13 (5.5) 7 (10.9) .001 
Hypertension, number (%) 103 (42.6) 78 (50.3) 17 (43.6) .31 120 (49.4) 111 (46.8) 42 (65.6) .027 
Diabetes, number (%) 29 (12.0) 20 (12.9) 7 (18.0) .54 19 (7.8) 27 (11.4) 9 (14.1) .21 
Osteoarthritis, number (%) 131 (54.1) 76 (49.0) 20 (51.3) .60 135 (55.6) 147 (62.0) 37 (57.8) .36 
Myocardial infarction, number (%) 11 (4.6) 11 (7.1) 5 (12.8) .11 5 (2.1) 11 (4.6) 3 (4.7) .21 
Angina, number (%) 14 (5.8) 13 (8.4) 3 (7.7) .51 0 (0) 11 (4.6) 4 (6.2) <.001 
Chronic obstructive pulmonary disease, number (%) 30 (12.4) 32 (20.6) 12 (30.8) .006 0 (0) 4 (1.7) 5 (7.8) <.001 
Creatinine clearance (mL/min), median (IQR) 70.5 (61.0–85.0) 68.5 (54.4–82.2) 57.8 (43.8–67.9) <.001 62.4 (50.4–73.2) 59.5 (47.0–70.7) 48.3 (37.3–61.8) <.001 
Serum PTH (ng/L), median (IQR) 19.2 (14.5–25.2) 24.2 (16.5–35.8) 23.6 (19.1–38.3) <.001 22.7 (16.2–32.0) 23.4 (15.4–35.8) 29.4 (20.2–44.1) .002 
Serum PTH >32.4 ng/L, number (%) 28 (11.6) 46 (29.7) 12 (30.8) <.001 58 (23.9) 71 (30.0) 27 (42.2) .014 
25(OH)D (nmol/L), median (IQR) 61.5 (41.7–78.4) 43.2 (30.0–69.6) 35.7 (23.5–48.4) <.001 36.2 (27.2–56.2) 34.2 (21.7–47.7) 26.4 (17.2–35.9) <.001 
25(OH)D <50.0 nmol/L, number (%) 96 (39.7) 97 (62.6) 30 (76.9) <.001 172 (70.8) 180 (76.0) 56 (87.5) .017 
Frailty criteria, number (%)         
    Weight loss 13 (8.7) 13 (35.1)  13 (5.7) 15 (26.3)  
    Exhaustion 29 (20.1) 20 (62.5)  84 (38.0) 35 (64.8)  
    Low energy expenditure 29 (18.7) 30 (76.9)  87 (36.7) 56 (87.5)  
    Slowness 66 (46.2) 33 (97.1)  68 (31.0) 53 (91.4)  
    Weakness 72 (46.4) 34 (87.2)  73 (30.8) 50 (79.4)  
Characteristic Men
 
Women
 
Robust (N = 242) Intermediate Frail (N = 151) Frail (N = 39) p Robust (N = 243) Intermediate Frail (N = 237) Frail (N = 64) p 
Age (yrs), mean (SD71.7 (5.5) 76.2 (7.1) 80.8 (7.8) <.001 72.9 (6.1) 76.1 (7.5) 82.3 (7.4) <.001 
Education (yrs), mean (SD6.5 (3.6) 5.9 (3.7) 4.6 (2.7) .003 5.1 (2.7) 4.5 (2.8) 4.2 (2.8) .025 
Season of blood collection, number (%)         
    Spring (March–May) 40 (16.5) 30 (19.4) 8 (20.5)  63 (25.9) 34 (14.4) 18 (28.1)  
    Summer (June–August) 27 (11.2) 20 (12.9) 9 (23.1)  18 (7.4) 27 (11.4) 9 (1.6)  
    Fall (September–November) 70 (28.9) 50 (32.3) 13 (33.3)  71 (35.7) 102 (43.0) 26 (40.6)  
    Winter (December–February) 105 (43.4) 55 (35.5) 9 (23.1) .17 91 (37.4) 74 (31.2) 11 (17.2) <.001 
Smoking (pack-years), mean (SD22.7 (23.5) 25.8 (24.2) 29.0 (31.2) .22 2.0 (6.4) 3.3 (9.0) 4.4 (15.4) .095 
Alcohol consumption (drinks/wk), mean (SD14.0 (13.4) 12.4 (14.2) 8.8 (9.0) .075 4.2 (5.7) 3.0 (4.7) 3.3 (5.5) .030 
BMI (kg/m2), mean (SD27.0 (3.4) 27.2 (3.3) 26.7 (2.8) .61 27.5 (4.3) 27.8 (4.5) 28.3 (6.1) .42 
MMSE score, mean (SD26.4 (2.7) 24.2 (5.3) 20.9 (7.6) <.001 25.2 (3.6) 23.8 (5.0) 21.6 (5.4) <.001 
Congestive heart failure, number (%) 8 (3.3) 11 (7.1) 7 (18.0) .003 1 (0.4) 17 (7.2) 11 (17.2) <.001 
Peripheral arterial disease, number (%) 15 (6.2) 15 (9.7) 6 (15.4) .11 3 (1.2) 13 (5.5) 7 (10.9) .001 
Hypertension, number (%) 103 (42.6) 78 (50.3) 17 (43.6) .31 120 (49.4) 111 (46.8) 42 (65.6) .027 
Diabetes, number (%) 29 (12.0) 20 (12.9) 7 (18.0) .54 19 (7.8) 27 (11.4) 9 (14.1) .21 
Osteoarthritis, number (%) 131 (54.1) 76 (49.0) 20 (51.3) .60 135 (55.6) 147 (62.0) 37 (57.8) .36 
Myocardial infarction, number (%) 11 (4.6) 11 (7.1) 5 (12.8) .11 5 (2.1) 11 (4.6) 3 (4.7) .21 
Angina, number (%) 14 (5.8) 13 (8.4) 3 (7.7) .51 0 (0) 11 (4.6) 4 (6.2) <.001 
Chronic obstructive pulmonary disease, number (%) 30 (12.4) 32 (20.6) 12 (30.8) .006 0 (0) 4 (1.7) 5 (7.8) <.001 
Creatinine clearance (mL/min), median (IQR) 70.5 (61.0–85.0) 68.5 (54.4–82.2) 57.8 (43.8–67.9) <.001 62.4 (50.4–73.2) 59.5 (47.0–70.7) 48.3 (37.3–61.8) <.001 
Serum PTH (ng/L), median (IQR) 19.2 (14.5–25.2) 24.2 (16.5–35.8) 23.6 (19.1–38.3) <.001 22.7 (16.2–32.0) 23.4 (15.4–35.8) 29.4 (20.2–44.1) .002 
Serum PTH >32.4 ng/L, number (%) 28 (11.6) 46 (29.7) 12 (30.8) <.001 58 (23.9) 71 (30.0) 27 (42.2) .014 
25(OH)D (nmol/L), median (IQR) 61.5 (41.7–78.4) 43.2 (30.0–69.6) 35.7 (23.5–48.4) <.001 36.2 (27.2–56.2) 34.2 (21.7–47.7) 26.4 (17.2–35.9) <.001 
25(OH)D <50.0 nmol/L, number (%) 96 (39.7) 97 (62.6) 30 (76.9) <.001 172 (70.8) 180 (76.0) 56 (87.5) .017 
Frailty criteria, number (%)         
    Weight loss 13 (8.7) 13 (35.1)  13 (5.7) 15 (26.3)  
    Exhaustion 29 (20.1) 20 (62.5)  84 (38.0) 35 (64.8)  
    Low energy expenditure 29 (18.7) 30 (76.9)  87 (36.7) 56 (87.5)  
    Slowness 66 (46.2) 33 (97.1)  68 (31.0) 53 (91.4)  
    Weakness 72 (46.4) 34 (87.2)  73 (30.8) 50 (79.4)  

Notes: The p values compare variables across frailty status, within gender. The χ2 or Fisher exact test for categorical variables; analysis of variance or Kruskal-Wallis test for continuous variables.SD = standard deviation, IQR = interquartile range. 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone; BMI = body mass index; MMSE = Mini-Mental State Examination.

Table 2 shows adjusted sex-specific associations of 25(OH)D with frailty. In men, ORs (95% confidence intervals [CIs]) of frailty and intermediate frailty, respectively, versus robustness [low vs high 25(OH)D] were 4.94 (1.80–13.61) and 2.52 (1.56–4.07). Men's OR for intermediate frailty, controlling for PTH, was attenuated by 16%, and high PTH was associated with intermediate frailty (OR = 2.42, 95% CI = 1.31–4.48). Among women, ORs (95% CI) of frailty and intermediate frailty, respectively, versus robustness were 1.43 (0.58–3.56) and 1.18 (0.75–1.88). OR for frailty for women was attenuated by 5% after controlling for PTH. Sex was an effect modifier of the association between low 25(OH)D and frailty (p = .041).

Table 2.

Multinomial Logistic Regression Assessing the Association of 25(OH)D With Frailty

Biomarkers Men
 
Women
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p 
Model 1*         
    25(OH)D (low vs high) 4.94 (1.80–13.61) .002 2.52 (1.56–4.07) <.001 1.43 (0.58–3.56) .44 1.18 (0.75–1.88) .47 
Model 2*         
    25(OH)D (low vs high) 4.13 (1.45–11.80) .008 2.13 (1.30–3.50) .003 1.36 (0.54–3.43) .52 1.17 (0.73–1.87) .51 
    PTH (high vs low) 2.40 (0.84–6.87) .10 2.42 (1.31–4.48) .005 1.37 (0.65–2.89) .41 1.16 (0.74–1.83) .52 
Biomarkers Men
 
Women
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p 
Model 1*         
    25(OH)D (low vs high) 4.94 (1.80–13.61) .002 2.52 (1.56–4.07) <.001 1.43 (0.58–3.56) .44 1.18 (0.75–1.88) .47 
Model 2*         
    25(OH)D (low vs high) 4.13 (1.45–11.80) .008 2.13 (1.30–3.50) .003 1.36 (0.54–3.43) .52 1.17 (0.73–1.87) .51 
    PTH (high vs low) 2.40 (0.84–6.87) .10 2.42 (1.31–4.48) .005 1.37 (0.65–2.89) .41 1.16 (0.74–1.83) .52 

Notes: 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone; OR = odds ratio; CI = confidence interval.

*

Model 1: 25(OH)D + covariates; Model 2: 25(OH)D + PTH + covariates; Low 25(OH)D: <50.0 nmol/L; High PTH: >32.4 ng/L. Covariates included age (years), education (years of schooling), blood collection season, smoking (pack-years), alcohol consumption (drinks/wk), cognitive function (Mini-Mental State Examination), comorbidities (congestive heart failure, peripheral arterial disease, hypertension, diabetes mellitus, osteoarthritis, myocardial infarction, angina, and chronic obstructive pulmonary disease), body mass index (kg/m2), and creatinine clearance (mL/min).

Covariate-adjusted analyses with continuous 25(OH)D showed that a 10 nmol/L difference in 25(OH)D (lower vs higher) was associated with ORs (95% CI) for frailty of 1.35 (1.10–1.66) and 1.04 (0.92–1.17) among men and women, respectively. Controlling for continuous PTH and covariates, ORs (95% CI) for frailty were 1.33 (1.08–1.64) and 1.02 (0.90–1.14) among men and women, respectively, (not shown).

Sex-specific associations of 25(OH)D with individual frailty criteria are shown in Table 3. After adjustment for covariates, men with low 25(OH)D had elevated odds of low energy expenditure (OR = 6.83, 95% CI = 2.62–17.82) and slowness (OR = 2.52, 95% CI = 1.39–4.59). After controlling for PTH, the OR for low energy expenditure in men increased by 9%, and men's OR for slowness decreased by 13%. Women with low 25(OH)D had elevated adjusted odds of low energy expenditure (OR = 2.28, 95% CI = 1.41–4.93), which remained unchanged after controlling for PTH.

Table 3.

Covariate-Adjusted Associations of 25(OH)D and PTH With Individual Frailty Criteria

 Weight Loss
 
Exhaustion
 
Low Energy Expenditure
 
Slowness
 
Weakness
 
 OR (95% CI) 
Men      
    Model 1*      
        25(OH)D (low vs high) 1.01 (0.38–2.73) 1.56 (0.78–3.12) 6.83 (2.62–17.82) 2.52 (1.39–4.59) 1.29 (0.75–2.23) 
p = .98 p = .20 p < .001 p = .002 p = .36 
    Model 2*      
        25(OH)D (low vs high) 1.08 (0.39–2.99) 1.19 (0.57–2.49) 7.47 (2.79–19.98) 2.20 (1.17–4.11) 1.04 (0.58–1.86) 
 p = .89 p = .64 p < .001 p = .014 p = .89 
    PTH (high vs low) 0.70 (0.20–2.48) 2.92 (1.34–6.39) 0.70 (0.29–1.70) 1.75 (0.88–3.45) 2.40 (1.27–4.55) 
p = .58 p = .007 p = .44 p = .11 p = .007 
Women      
    Model 1*      
        25(OH)D (low vs high) 1.45 (0.44–4.72) 0.76 (0.78–3.12) 2.28 (1.20–4.32) 1.23 (0.57–2.67) 0.79 (0.45–1.40) 
p = .54 p = .30 p = .012 p = .55 p = .42 
    Model 2*      
        25(OH)D (low vs high) 1.56 (0.47–5.14) 0.79 (0.47–1.34) 2.28 (1.19–4.37) 1.12 (0.59–2.14) 0.72 (0.40–1.29) 
p = .47 p = .39 p = .013 p = .72 p = .27 
    PTH (high vs low) 0.58 (0.19–1.75) 0.88 (0.52–1.48) 0.97 (0.57–1.63) 1.51 (0.88–2.58) 1.65 (0.99–2.76) 
p = .33 p = .62 p = .90 p = .13 p = .055 
 Weight Loss
 
Exhaustion
 
Low Energy Expenditure
 
Slowness
 
Weakness
 
 OR (95% CI) 
Men      
    Model 1*      
        25(OH)D (low vs high) 1.01 (0.38–2.73) 1.56 (0.78–3.12) 6.83 (2.62–17.82) 2.52 (1.39–4.59) 1.29 (0.75–2.23) 
p = .98 p = .20 p < .001 p = .002 p = .36 
    Model 2*      
        25(OH)D (low vs high) 1.08 (0.39–2.99) 1.19 (0.57–2.49) 7.47 (2.79–19.98) 2.20 (1.17–4.11) 1.04 (0.58–1.86) 
 p = .89 p = .64 p < .001 p = .014 p = .89 
    PTH (high vs low) 0.70 (0.20–2.48) 2.92 (1.34–6.39) 0.70 (0.29–1.70) 1.75 (0.88–3.45) 2.40 (1.27–4.55) 
p = .58 p = .007 p = .44 p = .11 p = .007 
Women      
    Model 1*      
        25(OH)D (low vs high) 1.45 (0.44–4.72) 0.76 (0.78–3.12) 2.28 (1.20–4.32) 1.23 (0.57–2.67) 0.79 (0.45–1.40) 
p = .54 p = .30 p = .012 p = .55 p = .42 
    Model 2*      
        25(OH)D (low vs high) 1.56 (0.47–5.14) 0.79 (0.47–1.34) 2.28 (1.19–4.37) 1.12 (0.59–2.14) 0.72 (0.40–1.29) 
p = .47 p = .39 p = .013 p = .72 p = .27 
    PTH (high vs low) 0.58 (0.19–1.75) 0.88 (0.52–1.48) 0.97 (0.57–1.63) 1.51 (0.88–2.58) 1.65 (0.99–2.76) 
p = .33 p = .62 p = .90 p = .13 p = .055 

Notes: 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone; OR = odds ratio; CI = confidence interval.

*

Model 1: 25(OH)D + covariates; Model 2: 25(OH)D + PTH + covariates; Low 25(OH)D: <50.0 nmol/L; High PTH: >32.4 ng/L. Covariates included age (years), education (years of schooling), blood collection season, smoking (pack-years), alcohol consumption (drinks/wk), cognitive function (Mini-Mental State Examination), comorbidities (congestive heart failure, peripheral arterial disease, hypertension, diabetes mellitus, osteoarthritis, myocardial infarction, angina, and chronic obstructive pulmonary disease), body mass index (kg/m2), and creatinine clearance (mL/min).

After refitting the model using continuous 25(OH)D, the ORs (95% CI) for each 10 nmol/L difference (lower vs higher) for men were 1.50 (1.23–1.83) and 1.10 (0.99–1.22) for low energy expenditure and slowness, respectively. For women, the OR for low energy expenditure was 1.07 (95% CI = 0.99–1.16). Weak associations between continuous 25(OH)D and other frailty criteria were found (p > .05) among both sexes (not shown).

Low 25(OH)D levels were most strongly associated with low energy expenditure, possibly due to reverse causality via limited sunlight exposure. To assess the relationship of low 25(OH)D with nonenergy expenditure aspects of frailty, multinomial regression was performed using a four-component frailty measure (where low energy expenditure is omitted) while controlling for covariates and low energy expenditure. Among men, 15 participants previously classified as intermediate frail were classified as robust using the four-component measure, and 19 men were reclassified from frail to intermediate frail. Similarly, 29 women previously classified as intermediate frail were classified as robust using the four components, and 40 women were reclassified from frail to intermediate frail. Regression results are shown in Table 4. Among men, the adjusted ORs (95% CI) for frailty and intermediate frailty were 2.18 (0.59–8.04) and 1.86 (1.15–3.00), respectively. After controlling for PTH, the frailty OR was attenuated by 32%. Among women, the adjusted ORs (95% CI) for frailty and intermediate frailty were 1.54 (0.31–7.58) and 0.84 (0.53–1.33). After controlling for PTH the frailty OR was attenuated by 6%. For both sexes, associations of PTH with frailty were stronger using the four-component frailty measure than when using the five-component measure.

Table 4.

Sensitivity Analysis: Multinomial Logistic Regression Assessing the Association of 25(OH)D and PTH With Four-Category Frailty (Excluding Low Energy Expenditure)

Biomarkers Men
 
Women
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p 
Model 1*         
    25(OH)D (low vs high) 2.18 (0.59–8.04) .24 1.86 (1.15–3.00) .011 1.54 (0.31–7.58) .60 0.84 (0.53–1.33) .45 
Model 2*         
    25(OH)D (low vs high) 1.47 (0.37–5.76) .58 1.55 (0.94–2.55) .086 1.44 (0.28–7.29) .66 0.79 (0.50–1.27) .33 
    PTH (high vs low) 5.63 (1.62–19.55) .006 2.53 (1.38–4.63) .003 1.47 (0.50–4.38) .49 1.44 (0.92–2.25) .11 
Biomarkers Men
 
Women
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
Frail vs Robust
 
Intermediate Frail vs Robust
 
OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p 
Model 1*         
    25(OH)D (low vs high) 2.18 (0.59–8.04) .24 1.86 (1.15–3.00) .011 1.54 (0.31–7.58) .60 0.84 (0.53–1.33) .45 
Model 2*         
    25(OH)D (low vs high) 1.47 (0.37–5.76) .58 1.55 (0.94–2.55) .086 1.44 (0.28–7.29) .66 0.79 (0.50–1.27) .33 
    PTH (high vs low) 5.63 (1.62–19.55) .006 2.53 (1.38–4.63) .003 1.47 (0.50–4.38) .49 1.44 (0.92–2.25) .11 

Notes: 25(OH)D = 25-hydroxyvitamin D; PTH = parathyroid hormone; OR = odds ratio; CI = confidence interval.

*

Model 1: 25(OH)D + covariates + low energy expenditure; Model 2: 25(OH)D + PTH + covariates + low energy expenditure; Low 25(OH)D: <50.0 nmol/L; High PTH: >32.4 ng/L. Covariates included age (years), education (years of schooling), blood collection season, smoking (pack-years), alcohol consumption (drinks/wk), cognitive function (Mini-Mental State Examination), comorbidities (congestive heart failure, peripheral arterial disease, hypertension, diabetes mellitus, osteoarthritis, myocardial infarction, angina, chronic obstructive pulmonary disease), body mass index (kg/m2), and creatinine clearance (mL/min).

DISCUSSION

In this study, we found strong independent associations of low 25(OH)D with frailty measured using the widely accepted five components in men and weak evidence for association in women. Low 25(OH)D was independently associated with low energy expenditure in both sexes and with slowness in men. The evidence of association using a modified four-component measure of frailty (excluding low energy expenditure) in men was much weaker. One explanation for this attenuation is the smaller number of men classified as frail using the four-component measure; however, we cannot rule out reverse causality of low physical activity with low 25(OH)D as an explanation of this result.

Sex-stratified analyses controlled for biological and social differences between men and women that may have obscured associations in studies that included both sexes in one analysis. These results were robust to modeling of 25(OH)D (continuous and dichotomous). The 25(OH)D results were attenuated when low energy expenditure was removed from the frailty measure and included as a covariate; however, the PTH results were strengthened.

Although low physical activity may lead to low 25(OH)D, low 25(OH)D may be associated with frailty through several other biological pathways, including effects on bone, muscle, and immunity. Effects of vitamin D on bone health are well known (13). Low 25(OH)D levels are associated, both directly and mediated through elevated PTH, with osteoporotic fractures and osteomalacia, a condition whereby new bone matrix is not properly mineralized (3). These conditions, along with poor balance, (22) may result in bodily pain (7) (eg, via progression of osteoarthritis) (23,24) and fear of falling, which may contribute to sedentariness (3,13,25). This pathway may also explain the strong associations between low 25(OH)D and low energy expenditure for both sexes found in this study.

Additionally, low 25(OH)D may affect frailty via effects on muscle strength. Vitamin D receptors (VDRs) are located in skeletal muscle cells (26), and low 25(OH)D may result in decreased muscle strength from both decreased muscle synthesis and altered contractile properties of muscle (27). Muscle protein synthesis is initiated by binding 1,25-(OH)2D to its nuclear receptor. The influence of 1,25-(OH)2D on calcium homeostasis is believed to influence contractile properties of muscle cells via both a VDR-mediated genomic pathway and a nongenomic rapid mechanism (28). Thus, the association between low 25(OH)D and frailty may be explained by associations of insufficient 25(OH)D with sarcopenia and muscle weakness because both are central to the frailty syndrome (1). Previous studies found associations of low 25(OH)D with low muscle strength (4), poor balance (22,29), and falls (6). Additionally, associations between elevated PTH and sarcopenia have been reported (4). In this study, a moderate association was found between low 25(OH)D and weakness in men.

A third pathway through which low 25(OH)D may affect frailty stems from hypothesized anti-inflammatory properties of vitamin D (13). The VDR has been identified on most immune system cells (30,31). Active metabolite 1,25-(OH)2D inhibits secretion by antigen-presenting cells of cytokine interleukin-12 (IL-12). Since IL-12 stimulates development of CD4+ helper T cells subtype 1 (TH1) and inhibits development of CD4+ helper T cells subtype 2 (TH2), 1,25-(OH)2D actions result in a shift away from a proinflammatory TH1-like profile toward an anti-inflammatory TH2-like profile (31). Low 25(OH)D has been associated with increased risk of TH1 cytokine-mediated autoimmune diseases including inflammatory bowel disease, rheumatoid arthritis, and type 1 diabetes mellitus (32).

Associations between serum markers of inflammation with functional decline, slowness, and frailty have been reported (33–39). Effects of low 25(OH)D on muscle may be mediated by proinflammatory cytokines, which are believed to adversely affect muscle strength and function (33,34,36,40). Recent research (9) hypothesized that low 25(OH)D may affect physical function via secondary hyperparathyroidism resulting in elevated cytokine levels, which have been linked to sarcopenia (39). In this study, an association between low 25(OH)D and slowness was found among men.

Sex differences in the relationship of 25(OH)D with frailty found in this study may have biologic and methodologic origins. Women in this study had lower 25(OH)D levels than men; thus, relationships were estimated for different ranges of 25(OH)D in men and women. In particular, only 29.2% of robust women compared with 60.3% of robust men had adequate 25(OH)D levels. Differences in the 25(OH)D distribution may be compounded by age differences (women were older), because higher 25(OH)D levels may be needed for older individuals to avoid secondary hyperparathyroidism (16). Lastly, comorbidities with differing prevalence between sexes were included in our analyses, which may be an important part of the pathway linking low 25(OH)D to frailty in women compared with men (32). To avoid extrapolation problems, the results should be interpreted as “sex-specific associations” of 25(OH)D with frailty for the 25(OH)D distributions encountered in this study rather than “sex comparisons.”

An advantage of this study is the large population-based sample with plasma 25(OH)D, the best clinical indicator of vitamin D body stores. The cross-sectional design precludes determining temporality of 25(OH)D and frailty; therefore, observed associations may be due to noncausal relationships (eg, unmeasured confounders). Although insufficient 25(OH)D leading to progression of the frailty syndrome is biologically plausible, frail persons are possibly more likely to limit their outdoor activities, leading to decreased sunlight exposure and low 25(OH)D. This explanation is consistent with the strong association between low 25(OH)D and low energy expenditure found in this study. Also, the temporal relationship between 25(OH)D and PTH cannot be established; thus, reverse causality cannot be excluded as an explanation. Further longitudinal studies should evaluate whether low 25(OH)D is associated with onset or progression of frailty and its individual dimensions in older, community-dwelling adults.

In conclusion, results suggest that, independent of PTH, men with low 25(OH)D have elevated odds of frailty and intermediate frailty compared with men with adequate 25(OH)D, with evidence suggesting different associations with individual frailty criteria. The odds of frailty among women with low 25(OH)D were moderately elevated. Also, elevated PTH was strongly associated with nonenergy expenditure aspects of frailty in men with a moderate association in women. Three biologic pathways (bone, muscle, and inflammation) may explain the association between low 25(OH)D and frailty, and the relative contribution of each pathway may depend on sex. These results exemplify the need for longitudinally examining the disablement process separately in men and women and for assessing the relationship between vitamin D and frailty in populations of women with a broader distribution of 25(OH)D levels.

FUNDING

This research was partly supported by an unrestricted grant from Procter and Gamble, SrL, Italy. Authors were supported by grants from the National Institutes of Health (NIH): R37 AG09901, 1 R01 AI6085901A1, K23 AG019161, Building Interdisciplinary Research Careers in Women's Health. Dr. Hicks, a BIRCWH scholar at the time of this work, was supported by NIH K12 HD43489.

The InCHIANTI study was supported by the Italian Ministry of Health and in part by the U.S. National Institute on Aging (contracts 263 MD 916413 and 263 MD 821336).

References

1.
Fried
LP
Tangen
CM
Walston
J
, et al.  . 
Frailty in older adults: evidence for a phenotype
J Gerontol A Biol Sci Med Sci.
 , 
2001
, vol. 
56
 (pg. 
M146
-
M156
)
2.
Bandeen-Roche
K
Xue
QL
Ferrucci
L
, et al.  . 
Phenotype of frailty: characterization in the women's health and aging studies
J Gerontol A Biol Sci Med Sci.
 , 
2006
, vol. 
61
 (pg. 
262
-
266
)
3.
Lips
P
Vitamin D deficiency and secondary hyperparathyroidism in the elderly: consequences for bone loss and fractures and therapeutic implications
Endocr Rev.
 , 
2001
, vol. 
22
 (pg. 
477
-
501
)
4.
Visser
M
Deeg
DJ
Lips
P
Low vitamin D and high parathyroid hormone levels as determinants of loss of muscle strength and muscle mass (sarcopenia): the Longitudinal Aging Study Amsterdam
J Clin Endocrinol Metab.
 , 
2003
, vol. 
88
 (pg. 
5766
-
5772
)
5.
Bischoff-Ferrari
HA
Willett
WC
Wong
JB
Giovannucci
E
Dietrich
T
Dawson-Hughes
B
Fracture prevention with vitamin D supplementation: a meta-analysis of randomized controlled trials
JAMA
 , 
2005
, vol. 
293
 (pg. 
2257
-
2264
)
6.
Snijder
MB
van Schoor
NM
Pluijm
SM
van Dam
RM
Visser
M
Lips
P
Vitamin D status in relation to one-year risk of recurrent falling in older men and women
J Clin Endocrinol Metab.
 , 
2006
, vol. 
91
 (pg. 
2980
-
2985
)
7.
Mascarenhas
R
Mobarhan
S
Hypovitaminosis D-induced pain
Nutr Rev.
 , 
2004
, vol. 
62
 (pg. 
354
-
359
)
8.
Bischoff-Ferrari
HA
Dietrich
T
Orav
EJ
, et al.  . 
Higher 25-hydroxyvitamin D concentrations are associated with better lower-extremity function in both active and inactive persons aged > or =60 y
Am J Clin Nutr.
 , 
2004
, vol. 
80
 (pg. 
752
-
758
)
9.
Houston
DK
Cesari
M
Ferrucci
L
, et al.  . 
Association between vitamin D status and physical performance: the InCHIANTI Study
J Gerontol A Biol Sci Med Sci.
 , 
2007
, vol. 
62
 (pg. 
440
-
446
)
10.
Puts
MT
Visser
M
Twisk
JW
Deeg
DJ
Lips
P
Endocrine and inflammatory markers as predictors of frailty
Clin Endocrinol (Oxf).
 , 
2005
, vol. 
63
 (pg. 
403
-
411
)
11.
Gloth
FM
III
Smith
CE
Hollis
BW
Tobin
JD
Functional improvement with vitamin D replenishment in a cohort of frail, vitamin D-deficient older people
J Am Geriatr Soc.
 , 
1995
, vol. 
43
 (pg. 
1269
-
1271
)
12.
Verhaar
HJ
Samson
MM
Jansen
PA
de Vreede
PL
Manten
JW
Duursma
SA
Muscle strength, functional mobility and vitamin D in older women
Aging (Milano)
 , 
2000
, vol. 
12
 (pg. 
455
-
460
)
13.
Lips
P
Vitamin D physiology
Prog Biophys Mol Biol.
 , 
2006
, vol. 
92
 (pg. 
4
-
8
)
14.
Thomas
MK
Lloyd-Jones
DM
Thadhani
RI
, et al.  . 
Hypovitaminosis D in medical inpatients
N Engl J Med.
 , 
1998
, vol. 
338
 (pg. 
777
-
783
)
15.
Puts
MT
Lips
P
Deeg
DJ
Sex differences in the risk of frailty for mortality independent of disability and chronic diseases
J Am Geriatr Soc.
 , 
2005
, vol. 
53
 (pg. 
40
-
47
)
16.
Maggio
D
Cherubini
A
Lauretani
F
, et al.  . 
25(OH)D Serum levels decline with age earlier in women than in men and less efficiently prevent compensatory hyperparathyroidism in older adults
J Gerontol A Biol Sci Med Sci.
 , 
2005
, vol. 
60
 (pg. 
1414
-
1419
)
17.
Ferrucci
L
Bandinelli
S
Benvenuti
E
, et al.  . 
Subsystems contributing to the decline in ability to walk: bridging the gap between epidemiology and geriatric practice in the InCHIANTI study
J Am Geriatr Soc.
 , 
2000
, vol. 
48
 (pg. 
1618
-
1625
)
18.
Ble
A
Cherubini
A
Volpato
S
, et al.  . 
Lower plasma vitamin E levels are associated with the frailty syndrome: the InCHIANTI study
J Gerontol A Biol Sci Med Sci.
 , 
2006
, vol. 
61
 (pg. 
278
-
283
)
19.
Fava
GA
Assessing depressive symptoms across cultures: Italian validation of the CES-D self-rating scale
J Clin Psychol.
 , 
1983
, vol. 
39
 (pg. 
249
-
251
)
20.
Pisani
P
Faggiano
F
Krogh
V
Palli
D
Vineis
P
Berrino
F
Relative validity and reproducibility of a food frequency dietary questionnaire for use in the Italian EPIC centres
Int J Epidemiol.
 , 
1997
, vol. 
26
 
(suppl 1)
(pg. 
S152
-
S160
)
21.
Folstein
MF
Folstein
SE
McHugh
PR
“Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician
J Psychiatr Res.
 , 
1975
, vol. 
12
 (pg. 
189
-
198
)
22.
Zamboni
M
Zoico
E
Tosoni
P
, et al.  . 
Relation between vitamin D, physical performance, and disability in elderly persons
J Gerontol A Biol Sci Med Sci.
 , 
2002
, vol. 
57
 (pg. 
M7
-
M11
)
23.
McAlindon
TE
Felson
DT
Zhang
Y
, et al.  . 
Relation of dietary intake and serum levels of vitamin D to progression of osteoarthritis of the knee among participants in the Framingham Study
Ann Intern Med.
 , 
1996
, vol. 
125
 (pg. 
353
-
359
)
24.
Zhang
Y
Hannan
MT
Chaisson
CE
, et al.  . 
Bone mineral density and risk of incident and progressive radiographic knee osteoarthritis in women: the Framingham Study
J Rheumatol.
 , 
2000
, vol. 
27
 (pg. 
1032
-
1037
)
25.
Tinetti
ME
Powell
L
Fear of falling and low self-efficacy: a case of dependence in elderly persons
J Gerontol.
 , 
1993
, vol. 
48
 (pg. 
35
-
38
)
26.
Simpson
RU
Thomas
GA
Arnold
AJ
Identification of 1,25-dihydroxyvitamin D3 receptors and activities in muscle
J Biol Chem.
 , 
1985
, vol. 
260
 (pg. 
8882
-
8891
)
27.
Wassner
SJ
Li
JB
Sperduto
A
Norman
ME
Vitamin D deficiency, hypocalcemia, and increased skeletal muscle degradation in rats
J Clin Invest.
 , 
1983
, vol. 
72
 (pg. 
102
-
112
)
28.
Boland
R
de Boland
AR
Marinissen
MJ
Santillan
G
Vazquez
G
Zanello
S
Avian muscle cells as targets for the secosteroid hormone 1,25-dihydroxy-vitamin D3
Mol Cell Endocrinol.
 , 
1995
, vol. 
114
 (pg. 
1
-
8
)
29.
Bischoff
HA
Stahelin
HB
Urscheler
N
, et al.  . 
Muscle strength in the elderly: its relation to vitamin D metabolites
Arch Phys Med Rehabil.
 , 
1999
, vol. 
80
 (pg. 
54
-
58
)
30.
Holick
MF
The vitamin D epidemic and its health consequences
J Nutr.
 , 
2005
, vol. 
135
 (pg. 
2739S
-
2748S
)
31.
van Etten
E
Mathieu
C
Immunoregulation by 1,25-dihydroxyvitamin D3: basic concepts
J Steroid Biochem Mol Biol.
 , 
2005
, vol. 
97
 (pg. 
93
-
101
)
32.
Peterlik
M
Cross
HS
Vitamin D and calcium deficits predispose for multiple chronic diseases
Eur J Clin Invest.
 , 
2005
, vol. 
35
 (pg. 
290
-
304
)
33.
Reuben
DB
Cheh
AI
Harris
TB
, et al.  . 
Peripheral blood markers of inflammation predict mortality and functional decline in high-functioning community-dwelling older persons
J Am Geriatr Soc.
 , 
2002
, vol. 
50
 (pg. 
638
-
644
)
34.
Cappola
AR
Xue
QL
Ferrucci
L
Guralnik
JM
Volpato
S
Fried
LP
Insulin-like growth factor I and interleukin-6 contribute synergistically to disability and mortality in older women
J Clin Endocrinol Metab.
 , 
2003
, vol. 
88
 (pg. 
2019
-
2025
)
35.
Payette
H
Roubenoff
R
Jacques
PF
, et al.  . 
Insulin-like growth factor-1 and interleukin 6 predict sarcopenia in very old community-living men and women: the Framingham Heart Study
J Am Geriatr Soc.
 , 
2003
, vol. 
51
 (pg. 
1237
-
1243
)
36.
Cesari
M
Penninx
BW
Lauretani
F
, et al.  . 
Hemoglobin levels and skeletal muscle: results from the InCHIANTI study
J Gerontol A Biol Sci Med Sci.
 , 
2004
, vol. 
59
 (pg. 
249
-
254
)
37.
Walston
J
McBurnie
MA
Newman
A
, et al.  . 
Frailty and activation of the inflammation and coagulation systems with and without clinical comorbidities: results from the Cardiovascular Health Study
Arch Intern Med.
 , 
2002
, vol. 
162
 (pg. 
2333
-
2341
)
38.
Schaap
LA
Pluijm
SM
Deeg
DJ
Visser
M
Inflammatory markers and loss of muscle mass (sarcopenia) and strength
Am J Med.
 , 
2006
, vol. 
119
 (pg. 
526.e9
-
526.e17
)
39.
Visser
M
Pahor
M
Taaffe
DR
, et al.  . 
Relationship of interleukin-6 and tumor necrosis factor-alpha with muscle mass and muscle strength in elderly men and women: the Health ABC Study
J Gerontol A Biol Sci Med Sci.
 , 
2002
, vol. 
57
 (pg. 
M326
-
M332
)
40.
Bautmans
I
Njemini
R
Lambert
M
Demanet
C
Mets
T
Circulating acute phase mediators and skeletal muscle performance in hospitalized geriatric patients
J Gerontol A Biol Sci Med Sci.
 , 
2005
, vol. 
60
 (pg. 
361
-
367
)

Author notes

Decision Editor: Darryl Wieland, PhD, MPH