Vitamin d deficiency among medical impatients

Vitamin D deficiency is common among elderly nursing home or housebound patients and is a major risk factor for bone loss. The prevalence of vitamin D deficiency among patients hospitalized for acute médical illness is unknown. To establish the prevalence of vitamin D deficiency among general medical inpatients, we evaluated 290 consecutive patients (ISO in winter and 140 ifl summer) admitted to genera) medical ward at Massachusetts General Hospital. Patiente were evaluated on admission for vitamin D intake, ultraviolet <uv) light exposure, and established risk factors for vitamin D deficiency. Fasting levels of 25-OH vitamin D (25D), PTH, and ionized calcium were measured. The mean age of the patients was 62 yrs, 93% were ambulatory, 21 % housebound, 8% from nursing homes, and 21 % used suppléments containing vitamin D. 56% of patients were vitamin D deficient (2SD < IS ng/ml) and 22% had severe vitamin D deficiency (2SD < 8 ng/ml). Vitamin D deficiency was présent in 63% of patients studied in winter and 49% of patients studied in summer. 25D levels were directly related to vitamin D intake and uv exposure. Mean (± SE) daily vitamin D intake was 7.5 ±0.4 ug/day. Of the patient who consumed less man the current RDA (5 ug/day), 69% were vitamin D deficient while 43% of those with intakes above the RDA were vitamin D deficient Serum PTH levels were inversely related to 25D levels and were increased in 37% of individuals with 25D levels < 8 ng/ml, 24% of those with 2SD levels of 8-15 ng/ml, and 12% of those with 25D levels > 15 ng/ml. Significant multivariate predictors of vitamin D deficiency were vitamin D intake, age, season, housebound status, and nee. These 5 variables gave a claanfication accuracy of 62%. Significant multivariate predictors of secondary hyperparamyroiditm were ionized calcium, serum creatinine, and age. These 3 variables gave a classification accuracy of 80%. Among a subset of 77 individuals < 65 yrs of age (mean 42), who were not housebound and were free of known risk factors for vitamin D deficiency or chronic debilitating illnesses, the prevalence of 25D deficiency was 41 %. We conclude that vitamin D deficiency is extremely common in this inpatient population, even in those subjects who appear to be relatively healthy and wftose reported dietary intake of vitamin D exceeds the RDA. These data suggest that consideration should be given to measuring 25D levels in all medical inpatients and/or routinely supplementing vitamin D intake in similar patient populations.