**4. Discussion**

The high prevalence of hypovitaminosis D found in the institutionalized elderly individuals of the present study was only significantly associated with the summer season, and this was found in the theoretical model involving several proximal and distal factors that determine vitamin D status. Moreover, the seasons of the year were considered as cross-sectional variables. The evaluation of the seasons of the year together with the assessment of the dietary intake of vitamin D (whether it was adequate or inadequate) was a relevant point of the present study, in that it was necessary to consider the two external sources of vitamin D obtained by the population.

Vitamin D is the only nutrient acquired from external sources, e.g., natural sunlight. It is estimated that an individual's highest daily vitamin D intake can be obtained by exposure to sunlight. However, in elderly individuals, the capacity of the cutaneous synthesis of vitamin D is undermined by the gradual loss of this physiological mechanism [9,17].

In epidemiological studies on vitamin D status, physical activity is often used as a proxy for time spent outdoors and, indirectly, for exposure to sunlight. Consequently, the elderly individuals who spend their lives at home indoors and those who spend less time doing outdoor physical activities have decreased levels of 25(OH)D [43].

Many of the elderly individuals were distributed by the summer season (60.1%), but although the region where the study was conducted has an annual average of UV indexes ranging from high to very high [44], it can be inferred that the studied population did not benefit from this environmental factor, as shown by the contradictory significant association between hypovitaminosis D and summer. We know that in tropical regions, the summer heat is extremely uncomfortable, causing most elderly individuals to avoid sunlight exposure. Similar reasons for this paradoxical finding were previously discussed, emphasizing di fferent weather conditions between areas or countries [23]. In addition, their failure to benefit from the summer season can also be explained because activities in the NHs follow a pattern—they start early in the day with personal hygiene and provision of food and medicines, and end with little or no sunlight exposure, because the times are inappropriate. Furthermore, factors such as health problems, logistics of the NH, and indoor leisure activities a ffect the access of the elderly individuals to sunlight exposure more often when compared to their non-institutionalized peers [19,45].

Other findings from our study, such as the high inadequacies of dietary vitamin D and calcium, physical inactivity, and multimorbidities may additionally have nullified the response to environmental conditions that would otherwise be favorable for vitamin D status. In regard to elderly Brazilian subjects, even in the summer, Caucasian males in the younges<sup>t</sup> group of individuals presented a significant increase in 25(OH)D levels, demonstrating that a seasonal variation in 25(OH)D concentrations was influenced by factors such as gender, ethnicity, and age [20].

Okan et al. noted that 25(OH)D levels and the UV index of elderly people living in NHs were significantly lower than those living in their own homes, though there was a positive correlation between UV index and 25(OH)D observed in both groups [46].

These data reinforce the need to discuss vitamin D status in the context of the institutionalization of elderly individuals, and demands a model that encompasses diversified variables, such as those used in our study.

Although aging has been reported as a factor that interferes with vitamin D metabolism [9,17], the 25(OH)D concentrations in elderly Brazilian individuals with regular sunlight exposure and physical activity practice were similar to the values found in young and healthy individuals. These results confirm that vitamin D metabolism is activated by sunlight exposure, despite the physiological limitations of elderly individuals [20]. Physical activity is considered to be one of the factors associated with higher 25(OH)D levels [14]. In light of this evidence, some benefits can be achieved with changes in the lifestyles of elderly individuals.

Most of the elderly individuals lived in nonprofit institutions and had lived 29% of their life as an elderly individual in an institutional setting, which was considered a distal factor of the study model and was associated with an increased risk of vitamin D insu fficiency and hypovitaminosis D when compared to elderly individuals living in the community [19]. In the same study, the prevalences of hypovitaminosis D in institutionalized elderly individuals and outpatients from a city in southeastern Brazil were 71.2% and 43.8%, respectively. This result is similar to that found in the present study. In this regard, confounding factors arising from disease burden in elderly individuals may also account for the high prevalence of vitamin D insu fficiency. In a previous study, frail elderly individuals who had been recently institutionalized showed a 41% prevalence of vitamin D deficiency and a 45% prevalence of vitamin D insu fficiency [47].

In our study, the population profiles and the infrastructures of the NHs were similar. Therefore, the studied elderly individuals did not show significant di fferences, whether they were in nonprofit public or for-profit private institutions. As shown in Table 3, the high prevalence of hypovitaminosis D in for-profit private institutions demonstrates that the institutionalization, by itself, can be addressed as a determinant of health problems in NHs. Furthermore, the study population lives in the Northeast of Brazil, a region with high social inequalities that may a ffect genetic distinctions in comparison to other areas of the country. A study developed in the same region found evidence that longevity in this region was characterized by the presence of longer telomeres in women associated with a low level of education [48].

All these factors may partly explain the findings reported in this study. However, the high prevalence of vitamin D insu fficiency, per se, in the context of low dietary intake of vitamin D and calcium as well as physical inactivity, represents signs of deterioration in vitamin D status, thus a ffecting the health of institutionalized elderly individuals.

Other points should be addressed in the field of aging. This study focuses on this particular generation of long-living individuals, and according to the theoretical model proposed, age is a proximal factor of interference in 25(OH)D concentrations. In this regard, there is increasing evidence that aging can proceed at variable rates and be regulated by vitamin D. Normal concentrations of vitamin D are capable of maintaining these processes at their regular low rates, and this slows down the aging process and also helps to prevent the onset of some age-related diseases. When vitamin D is deficient, there is an increase in the activities of these aging processes, which not only accelerates the rate of aging, but also creates conditions that initiate the onset of age-related diseases [5].

The present study showed limitations regarding its temporal approach (cross-sectional), hindering the establishment of a causal relationship, and also due to lack of evaluation of sun exposure. Future studies should focus on reproducibility, involving a protocol with longitudinal observation and comparison among institutionalized and non-institutionalized elderly individuals. Furthermore, as shown by the benefits found in institutionalized elderly individuals in Finland [49] who received vitamin D food fortification, the literature reinforces the need for special attention in health care for elderly people, with a focus on modifiable risk factors.
