*2.1. Study Population*

The study was carried out with 304 elderly individuals living in nine Nursing Homes (NH) in the city of Natal (Rio Grande do Norte state), which is in the northeast of Brazil. All the individuals included in the study were at least 60 years old and living in these NHs. In Brazil, the Statute of the Elderly [26] considers elderly individuals to be aged 60 years or older. Elderly individuals who had marked physical and mental impairments, as well as difficult venous access, were excluded from the study. After exclusion, the initial sample of 304 elderly individuals was narrowed down to 153 participants (Figure 1).

**Figure 1.** Flowchart of the study participants.

The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Research Ethics Committee of the Federal University of Rio Grande do Norte (#263/11; CAAE 0290.0.051.000-11). All participants or their guardians provided written informed consent before they participated in the study.

The sample size calculation for the cross-sectional studies was defined on the basis of the mean and standard deviation of the concentration of 24.0 (7.6) ng/mL of 25(OH)D, corresponding to 10 random subjects in the sample of this study, and to the power of the test of 95%, resulting in 154 elderly individuals. Thus, this study obtained a response rate of 99.3%. For association studies, the sample size would be 156 elderly individuals, considering the population size of 304 elderly individuals, a hypothetical frequency of vitamin D deficiency of 71.2% [19], a confidence limit of 5%, and a design effect equal to 1 [27].

#### *2.2. Theoretical Model of the Study*

The theoretical model of this study was developed according to determinant factors of vitamin D status. Serum 25(OH)D concentration was defined as the dependent variable, and the independent variables were those likely to influence vitamin D status: Age, sex, skin color, body mass index (BMI), physical activity practice, mobility, diet, use of vitamin D supplements, and use of antiepileptic medications. The sociodemographic, nutritional, lifestyle, and biological variables were included in the "proximal" layers. Type of NH, age-adjusted time of institutionalization, and level of education were components of the "distal" layer. The season of the year was considered as a cross-sectional influence because it has a direct relationship with 25(OH)D concentrations (Figure 2). This study also addressed the relationship between the dependent variable and possible clinical implications in cases of vitamin D-deficient elderly individuals—multimorbidity, falls, sarcopenia, depression, cognition, and functional status.

**Figure 2.** Theoretical model of the determinants of vitamin D status in institutionalized elderly individuals.

Age-adjusted time of institutionalization was calculated to determine how much time after the advanced age period (as a percentage, defined as the time after turning 60 years of age) elderly individuals remained institutionalized, according to a previous study [28].

Weight, height, and waist and calf circumference were measured to determine the anthropometric variables. Skin color was obtained by self-classification among the five categories adopted by the Brazilian Institute of Geography and Statistics (IBGE): Black, mixed race ("pardo" in official Portuguese), Caucasian, Asian, and indigenous (Native Brazilian) [29]. Body weight was assessed with a Balmak® electronic scale (Santa Barbara do Oeste, Brazil) with a capacity of 300 kg and an accuracy of 50 g. In elderly individuals with disabilities or permanent mobility restrictions, weight was measured with a Seca® 985 bed scale (Bolton, England). Height was defined using the average of the two measurements and a portable Caumaq® stadiometer (Cachoeira do Sul, Brazil). For physically impaired elderly individuals, height was estimated using the equation of Chumlea et al. (1987) [30]. BMI was classified while considering the cutoff points proposed by Lipschitz [31].

Calf perimeter values less than 31 cm were indicative of loss of muscle mass in elderly individuals. Results for waist perimeter greater than 102 cm in males and 88 cm in females were considered to indicate a higher risk of diseases [32].

Physical activity practice was determined through information provided by caregivers, categorized into "yes" or "no" answers. Information on diseases, multimorbidity, polypharmacy, use of antiepileptic drugs, and vitamin D supplementation was collected from the medical records of the NHs. Multimorbidity was considered as the presence of two or more diseases with clinical diagnoses included in medical records. Polypharmacy was characterized for those individuals who took five or more medications daily [33].

Data on the mobility variable were collected from the caregivers of the elderly individuals using a Barthel scale [34]. Sarcopenia was measured according to previously defined criteria [35], based on discrimination of reduced muscle mass concentrations (calf circumference assessment) associated with a reduction in strength (grip test) and/or functional status. Functional status was assessed according to the scale set by Katz [36]. Cognitive function was assessed according to Pfeiffer's Short Portable Mental Status Questionnaire, which evaluates short- and long-term memory, orientation, information on everyday facts, and mathematical reasoning. Individuals were classified into four categories: intact mental function (0–2 errors), mild cognitive impairment (3–4 errors), moderate (5–7 errors), and severe (8–10 errors) [37,38].

The elderly subjects were categorized by seasons. The season used was the date corresponding to 30 days prior to blood collection. This was to take into account the half-life of 25(OH)D [39]. The Laboratory of Environmental and Tropical Variables of the Federal University of Rio Grande do Norte provided UV values for each season of the year [39].

#### *2.3. Analysis of 25(OH)D Serum Concentration*

Blood samples after overnight fasting (8–12 h) were collected by standard venipuncture. Serum concentrations of 25(OH)D were measured using the chemiluminescent Liaison® test from kit DiaSorin® (Saluggia, Italy). Serum 25(OH)D values equal to or below 29 ng/mL (72 nmol/L) were considered as "insufficient", while values between 30–60 ng/mL (75–150 nmol/L) were classified as "sufficient" [8]. In the present study, the name "hypovitaminosis D" was used for elderly individuals with a 25(OH)D value equal to or below 29 ng/mL.
