Excessive abdominal adiposity and body fat are associated with lower serum vitamin D levels : A population-based study

Objective To estimate the prevalence of 25-hydroxyvitamin D defi ciency and to analyze factors associated with lower serum vitamin levels in Brazilian adults. Methods A cross-sectional, population-based study consisted of 626 adult individuals of both sexes living in the urban area of Viçosa, Minas Gerais. The dependent variable used was the serum level of 25-hydroxyvitamin D and the independent variables were sociodemographic, anthropometric and body composition variables. The associations among the variables were verifi ed using simple and multiple linear regression models, considering alpha lower than 0.05 for the input in the fi nal model.

Excessive abdominal adiposity and body fat are associated with lower serum vitamin D levels: A population-based study Excesso de adiposidade abdominal e de gordura corporal se associa à menor concentração sérica de vitamina D: um estudo de base populacional

I N T R O D U C T I O N
It has long been believed that vitamin D played a role in bone mineral metabolism.Currently, it can be considered a steroid hormone that acts on metabolism and cell proliferation [1].Vitamin D deficiency has also been associated with some diseases, including cardiovascular diseases [2], obesity and diabetes [3][4][5][6], which are related to low bone mass (osteopenia and osteoporosis) and related factors (bone fractures) [7].
Different factors, such as age, sex, ethnicity, diet and sun exposure, may promote changes in serum vitamin D levels [8], contributing individually or collectively to the deficiency of this nutrient.Thus, adequate attention should be given to each of these factors to promote appropriate interventions.
It has been observed that 25-hydroxyvitamin D (25(OH)D) deficiency/insufficiency is a reality in different countries around the world, both in those located in low-latitude regions and in those that have adopted food fortification http://dx.doi.org/10.1590/1678-98652018000600002policies for several years [9,10].Concomitantly, an increase in the prevalence of overweight/ obesity suggests that changes in 25(OH)D levels may be related to body fat.Many theories have emerged with the purpose of explaining such an association, but all of them require further investigation as there are still many controversies to be addressed.
In the literature, some studies have shown an inverse association between serum 25(OH)D levels and body composition [11,12], whereas there is also a report of non-association [13].Thus, careful analysis of the methodological procedures used in the existing studies can help explain the different results, as well as the possible biases.
In the specific case of Brazil, although it is a tropical country with a year-round incidence of solar radiation, a significant prevalence of 25(OH)D deficiency has also observed [14,15].Recent evidence confirms similar findings in other populations around the world [16,17] that rekindles the question of which factors could be associated with a lower serum 25(OH)D levels in the adult population.
The growing concern regarding the lack of vitamin D as a universal problem, which is often not recognized and adequately treated [18,19], had led to an increase in the literature addressing the issue over the last decade.However, in Brazil, population-based studies using the adult population that aim to show the relationship between vitamin D and excess weight are scarce, as well as studies that point out mechanisms that may best explain this association [20,21].Given this shortcoming, our study aims to estimate the prevalence of 25(OH) D deficiency and analyze the factors associated with the low serum vitamin D levels in Brazilian adults.

M E T H O D S
Data from this cross-sectional populationbased study were obtained from the Health and Food Study (HFS) conducted between 2012 and 2014 in the city of Viçosa, Minas Gerais, Brazil.
The study population consisted of adults, aged 20-59 years, of both sexes, living in the urban area of the city.The sample calculation was determined by using the formula for prevalence estimates, considering the total number of individuals aged 20-59 years living in the urban area of Viçosa (43,431 people) [22], estimated prevalence of 77.4% [21], 95.0% confidence level, sample error of 5.0%, and estimated deff (design effect, cluster sampling) of 1.9.There was an increase of 10.0% for losses and refusals and 10.0% for controlling for confounding factors [23].With the aid of public domain program Epi Info, version 3.5.2(Centers for Disease Control and Prevention, Atlanta, Georgia, United States) [24], calculations indicated a minimum sample size (n) of 617 participants.
Data collection occurred between 2012-2014 and a double-stage cluster sampling was carried out.The first step of the research established the census area and the second step determined the households.The research consisted of the following steps: household visits to apply a structured questionnaire, laboratory blood collection carried out by a trained professional nurse, and assessment of anthropometric measures [23].
The following individuals were excluded from the study: pregnant women, postpartum women, bedridden individuals or those whose measurements could not be obtained, individuals with cognitive/intellectual difficulties, and those who had difficulty answering the questionnaire.
The dependent variable was the serum 25(OH)D level (ng/ml), which was evaluated by chemiluminescence [25] using the Architect 25(OH)D kit and Architect/Abbott (São Paulo, Brazil) equipment.Blood samples were collected using a vacuum-assisted blood collection system and disposable material after 12 hours of fasting.The status of 25(OH)D was determined according to the following reference values: sufficient (≥30.0ng/ml),insufficient (21.0ng/ml to 29.9ng/ml) and deficient (≤20.9ng/ml)[14,26].The season of the year when blood collection was performed was also evaluated and categorized as: winter, autumn, spring and summer.
The sociodemographic variables were as follows: sex (categorized as men and women), age (completed years and classified in ten-year periods [20 to 29, 30 to 39, 40 to 49 and 50 to 59 years]), educational level (full years of schooling, classified as 0-4, 5-8, 9-11 and ≥12 years of study), and marital status classified as unmarried (single/divorced/separated/widowed) or with a partner (married/with a partner).
The excess of abdominal adiposity was evaluated by measuring the abdominal circumference using an inelastic 2-meter measuring tape (Sanny ® , São Paulo, Brazil).The measurement was performed at the midpoint between the iliac crest and the last rib.Measurements were performed in triplicate by a single examiner and the measurement means were used in the analyses.The cut-off point used to determine the prevalence of excess abdominal adiposity was based on abdominal circumference (AC) according to the following reference values: men CA≥90cm and women CA≥80cm [27].
Body fat mass was determined using bioimpedance (A-310, Biodynamics Corporation, Shoreline, Washington, United States).To carry out the examination, all the participants of the study were instructed to follow these procedures: fast before examination; no consumption of alcoholic beverages or practice of vigorous exercises 24 hours prior to the examination; urinate at least 30 minutes before the test [28].The measurements were performed with the individual in the supine position, wearing light clothing and free of any metal objects.Four electrodes were placed on the right side, two on the hands and two on the feet.Excess body fat mass was established in accordance with Lohman's proposal for adult individuals [29].
Descriptive analysis of variables was analyzed by means of relative frequency.Student's t-test and one-way Analysis of Variance (ANOVA) were used to analyze the difference between the mean 25(OH)D levels (ng/ml) and sociodemographic variables.Linear regression models were used to verify the association between the independent variables and serum 25(OH)D levels.The variables that presented p<0.20

R E S U L T S
Of the 626 participants in the survey, most were women (56.21%), 39.23% of those surveyed were between the ages of 20-29 years, 52.07%reported more than twelve years of schooling, and most self-reported being nonwhite (58.70%).As for the nutritional status, 55.92% presented excess abdominal fat and 43.43% presented excess body fat (Table 1).
Vitamin D (25(OH)D) deficiency (<20ng/ ml) was present in 14.4% of the participants and 42.0% of adults had 25(OH)D insufficiency (20-29ng/ml).Excess abdominal fat was higher in participants with 25(OH)D sufficiency (Figura 1).Body fat was higher in men with vitamin D insufficiency (Figure 2).Mean serum 25(OH)D level was 30.34ng/ ml and standard deviation was 9.85ng/ml, which were higher in males (p<0.001),among younger individuals (p=0.002),individuals who reported higher schooling (p=0.016),individuals whose blood was collected in the summer (p<0.001),  and in adults without excess abdominal or body adiposity (p<0.001)(Table 2).
Table 3 shows the association between explanatory variables and serum 25(OH)D levels.Negative associations were found between serum 25(OH)D level and age, showing that the older the age group, the lower the serum vitamin D level (p<0.001).Increased years of schooling also showed an inverse association with 25(OH) D level.As for the season of the year when blood was collected, this variable was positively associated with the increase in the serum vitamin level.We found that the serum 25(OH) D level was higher in individuals whose blood was collected in the summer than in individuals whose blood was collected in other seasons of the year.The serum 25(OH)D level was inversely associated with excess body adiposity (p=0.028) and excess abdominal adiposity (p=0.023).

D I S C U S S I O N
In our cross-sectional study with Brazilian adults, 25(OH)D insufficiency was high as it was present in 42% of the participants.According to Holick [30] and James [31], vitamin D deficiency is a public health problem in many countries, affecting approximately 1 billion people worldwide.In several regions of Brazil there is a high prevalence of vitamin deficiency in different age groups and both sexes [32]; however, we do not yet have a national study that describes the real scenario.In addition, the mean serum 25(OH)D level was higher among men, younger individuals who reported a higher educational level, and among individuals with a higher socioeconomic status.
Several risk factors are associated with hypovitaminosis D in developing countries such as Brazil [30].Among these factors, women have been described in epidemiological studies.Recent research has also shown that both intake and vitamin D synthesis are greater in men than in women [33][34][35].In addition, women spend less time exposed to the sun, practice less physical activity and have greater body adiposity.
Hypovitaminosis D is also observed in older individuals; thus, younger individuals have higher 25(OH)D levels.A study conducted in São Paulo with 591 volunteers of both sexes showed that the younger the research participant, the higher the serum vitamin D level [36].Unger et al. [37] identified this same association in relation to age when evaluating the population  of a Brazilian capital.The possible explanation for these results may be the pattern of different vitamin D intake according to the age groups [38] as well as non-supplementation or food fortification that are necessary for older individuals.In addition to this association, we can infer that older individuals spend less time exposed to the sun, which consequently leads to vitamin D deficiency, since sun exposure is the essential for vitamin D synthesis [35].
Lower educational levels may also be associated with hypovitaminosis D, which may be related to the poor access to information and good health behaviors, such as sun exposure and nutrition.These results corroborate a study by Daly et al. [39] who identified that a higher level of education is directly associated with sufficient vitamin D status.No studies with Brazilian adults have been identified to establish a relation between sun exposure and educational levels.Regarding serum vitamin D levels, according to the season the blood samples were collected, we found that individuals whose blood was collected in the summer presented higher vitamin D levels when compared with the individuals whose blood was collected in other seasons of the year.According to Webb et al. [40], solar radiation reaching the Earth, especially the UVB spectrum, significantly influences cutaneous vitamin D synthesis, so vitamin D synthesis is affected during the seasons of the year when sunlight exposure is lower, which leads to hypovitaminosis.
In our study, it was found that hypovitaminosis D was associated with excess body and abdominal adiposity.Similar studies have described the association between body fat and vitamin D status, which may be related to hormonal and nutritional factors [38].Clinical studies point to vitamin D deficiency as one of the factors that triggers the accumulation of http://dx.doi.org/10.1590/1678-98652018000600002body fat through reactions that inhibit lipolysis and stimulate lipogenesis [41,42].According to Wortsman et al. [43] and Silva et al. [44], evidence has suggested that the presence of receptors in adipose tissues results in vitamin D retention in adipocytes and thus the vitamin bioavailability in the tissues is reduced.Khan et al. [45] investigated the association between serum vitamin D level and adiposity in 3020 adult Afro-American subjects and suggests that Vitamin D Receptor (VDR) polymorphisms increase susceptibility to excess body and abdominal adiposity.In addition, hypovitaminosis D in obese individuals may be related to less physical activity in open-air settings and, consequently, these individuals spend less time exposed to the sun [46].
Although this is a cross-sectional study and since these associations cannot be interpreted as a causal relationship, the results may contribute to a better understanding of vitamin D metabolism in adults.In addition, the limitation of the study was that factors related to sun exposure, such as professional activity, and use of sunscreen were not investigated.The positive factor is that it is a population-based study that associates vitamin D with sociodemographic, anthropometric and body composition factors.

C O N C L U S I O N
Our results suggest that excess body and abdominal adiposity are strong predictors of changes in serum vitamin D levels; thus, public policies for prevention and treatment are essential.In addition, large clinical studies should be conducted to determine the major risk factors for hypovitaminosis D.

Figure 2 .
Figure 2. Mean values of body adiposity and Vitamin D status according to sex in adults.Health and Food Study (HFS).Viçosa (MG), Brazil, 2012-2014.
C O N T R I B U T O R SKJ SEGHETO, DCG SILVA, FG FERREIRA and GZ LONGO collaborated with design, analysis and interpretation of data, revision and approval of the final version of the manuscript.MR CARVALHO contributed with revision and approval of the final version of the manuscript.
in the bivariate analysis were included in the multiple model and only variables at p<0.05 remained in the model.The Stata software (Stata Corportation, College Station, Texas, United States) program, version 13.1, was used.The study was approved by the Research Ethics Committee of the Universidade Federal de Viçosa (UVF, Federal University of Viçosa) under report No 008/2012/CEPH.The interviewers read consent term to the interviewees who consented to participate in the study.
Figure 1.Mean values of abdominal adiposity and vitamin D status according to sex in adults.Health and Food Study (HFS).Viçosa (MG), Brazil, 2012-2014.