Senior citizens in nursing homes often have limited access to direct sunlight because of mobility issues. Nevertheless, Thailand is located near the equator where sufficient sunlight is available to provide adequate ultraviolet B exposure all year round. Elderly Thai nursing home residents were expected to have less prevalence of vitamin D insufficiency compared with people in countries located at high latitudes. However, mean serum level of 25(OH)D in the elderly nursing home residents in this study was 65

nmol/L and 61.3% of them had vitamin D insufficiency (≤ 70

nmol/l). The level of 25(OH)D in the present study was lower than the level in a population of younger normal health-conscious Thai women (73–129

nmol/L)
[
17]. In addition, there were no difference of the mean serum 25(OH)D concentration between those women taking and those not taking vitamin D supplement. This may result from the limited number of vitamin D supplement subjects in the group. The mean serum 25(OH)D level of elderly Thai nursing home residents was still higher than populations in countries where fortification of milk and vitamin D supplements is common such as in the United States (mean 25(OH)D level

=

30

nmol/l)
[
18]. Elderly people in Japan where consumption of oily fish high in vitamin D is part of the usual diet, had the mean 25(OH)D as low as 29.9

nmol/l
[
19]. This evidence supports the role of sunlight exposure in elderly people to improve vitamin D status.
Interestingly, dietary calcium intake did not correlate with either lumbar spine or femoral BMD in this population. This may be explained by the fact that all residents are provided with the same food each meal. Mean calcium intake was 322 ± 158

mg/day, which was higher than previously report in Thai farmers (236 ± 188

mg/day)
[
20] and was similar to other populations of Thai urban postmenopausal women (348.9 ±12.9

mg/day)
[
21]. It is much lower than that reported in a Caucasian population (670 ± 258

mg/day)
[
22]. Lactose intolerance can be a barrier to milk consumption among Asians leading to the consumption of fewer calcium-containing foods than Caucasian women
[
23]. The high normal level of serum PTH in this study (mean 4.3, normal range 1.1 - 5.8 pmol/L) can be partially explained by the low calcium intake and less calcium intestinal absorption in the elderly. Moreover, a significant positive relationship between serum PTH and CTx levels was found. The US National Institute of Health recommends the daily calcium requirement to be as high as 1,500

mg calcium in postmenopausal women not using hormone replacement therapy and in both men and women aged over 65

years
[
24]. However, the addition of 500

mg/day of elemental calcium would be optimal to achieve maximum calcium retention in Asians as demonstrated by a balanced study in Japanese postmenopausal women
[
25]. Seniors living in nursing homes should be encouraged to increase outdoor activities and sunlight exposure with additional consumption of calcium-rich food or calcium and vitamin D supplement to reduce the risk of osteoporotic fracture.
Serum 25(OH)D levels had an inverse relationship with age in previous reports
[
26,
27]. However, age-related decline in serum 25(OH)D was not consistent in other reports or in the present study
[
19,
28,
29]. In addition, an inverse association between serum 25(OH)D and PTH levels was not observed in our study. This might be explained by the limited number of subjects with low vitamin D insufficiency (N

=

36; 38.7%). Furthermore, there was no association between vitamin D or PTH levels and BMD. This result is similar to previous reports on Asian women
[
30,
31]. On the other hand, European and North American studies have shown a correlation between serum 25(OH)D concentration and vertebral or femoral BMD
[
32,
33]. From a previous study, positive relationship between serum 25(OH)D and femoral neck BMD will only be significant when serum 25(OH)D is lower than 30

nmol/l
[
34]. In our population there was only one subject with 25(OH)D level less than 30

nmol/L.
Based on the WHO classification, the prevalence of osteoporosis among nursing home residents was 47.3%. Percentage of osteoporosis at femoral neck was higher than at the lumbar spine. This prevalence was relatively lower than other studies in a Western population that varied from 55% to 85.8%
[
35,
36]. Besides genetic differences and life style factors, a selection bias might lead to lower prevalence of osteoporosis because the eligibility criteria in this study excluded all subjects who were not healthy.
Factors associated with low BMD were identified in this study. Both energy expenditure and serum CTx were correlated with lumbar and femoral neck BMD. The positive effect of high physical activity to increased BMD and reduced osteoporotic fracture rate has been shown in various studies for many different ethnic groups
[
37]. CTx is a bone resorption marker, and has been shown to be a convenient tool for monitoring BMD after treatment
[
38]. In a previous report, serum CTx level was a better indicator than other bone markers, providing an earlier indication of response
[
39]. BMD measurement is relatively expensive and may not be accessible to all individuals. Serum CTx level could use as an alternative screening and monitoring method for individuals who may be suffering from osteoporosis. This strategy was proposed because it was more convenient for nursing home residents who cannot easily travel to the hospitals where BMD measurement facilities are available. A CTx cut-off value for this nursing home population was found to be 0.237

ng/ml with a sensitivity of 81%. This implication applies to elderly persons who have a serum CTx level higher than 0.237

ng/ml; they should be monitored more closely for the presence of osteoporosis such as arranging for the more expensive BMD scan. However, the cost-effectiveness should be evaluated in the future.
Limitations
A number of limitations exist with respect to the present study. Firstly, this is a cross-sectional study; the correlations cannot imply the causation relationships between parameters and low BMD. Moreover, the study could not include all the residents who resided in Thai nursing homes. The data represent only subjects who met the eligibility criteria and agreed to participate. Urine samples may not be adequately collected due to the inconvenience of the procedure. However, this limitation had been corrected by expressing data as a ratio relative to urine creatinine. Furthermore, energy expenditure of the elderly subjects was obviously higher than the energy intake. Energy expenditure per day was calculated on the basis of energy requirement for each activity expressed in terms of METs. The high energy expenditure result might be overestimated due to the formula of the MET, which is more suitable for young adults whereas the subjects in this study are elderly. This could be explained from the previous study that overestimation of energy expenditure was greater with older age
[
40]. Lastly, there were 9 subjects who had been taking active form of vitamin D supplements for about 3

months. They were not excluded because their 25(OH)D levels were not different from the rest.