Abstract

Background
Evidence on the association between single nucleotide polymorphisms (SNPs) in the vitamin D receptor (VDR) and depressive symptoms is inconclusive.

Objectives
The primary aim of the study was to investigate the association between SNPs in the VDR gene and depressive symptoms.

Methods
In a sample of older adults from the Longitudinal Ageing Study Amsterdam (n = 922), depressive symptoms were assessed using the Centre for Epidemiological Studies Depression scale (CES-D scale) at baseline and after 3, 6, and 10 y of follow-up. Blood samples for SNP and serum 25-hydroxyvitamin D3 (25(OH)D3) determination were obtained at baseline. The association between 13 SNPs in the VDR gene and the course of depressive symptoms were evaluated using linear mixed models. The interaction between SNPs and serum 25(OH)D3 in relation to depressive symptoms was evaluated using multiple linear regression.

Results
No SNPs were associated with the course of depressive symptoms. Significant interactions between serum 25(OH)D3 and SNPs in the VDR gene were found. Stratified analysis revealed that within the GG genotype strata, 10 nmol/L higher serum 25(OH)D3 was associated with 0.27 (95% CI: −۰٫۵۰, −۰٫۰۴) and 0.23 (95% CI: −۰٫۴۸, ۰٫۰۲) lower scores on the CES-D scale for Cdx-2 and 1b-G-886A, respectively. This association was not found in persons having the GA or AA genotype.

Conclusions
No SNPs is associated with the course of depressive symptoms. Stratified analysis shows that the effect of serum 25(OH)D3 concentrations on depressive symptoms is different among genotypes of Cdx-2 and 1b-G-886A. Future research should elucidate on the function of Cdx-2 and 1b-G-886A to describe their effect.

Introduction

Depressive symptoms follow a U-shaped relationship with age, increasing again after the age of 60–۶۵ y [1]. This is, however, not solely due to aging and varies widely between people [2]. A higher medical burden [3], aging-related anxiety [4], and nonhealth-related events [5] have been associated with depressive symptoms in older adults. The prevalence of depressive symptoms in older adults is estimated at 17% [6], but they are often overlooked [7] due to the overlap with symptoms, among others [6,8]. The burden will continue to rise due to the aging population and increasing life expectancy [9], thereby increasing years lived with disease [10] and societal costs [11].

Different factors (e.g., biological, social, and psychological) play a role in the onset and course of depressive symptoms [[12], [13], [14], [15]]. Vitamin D – a modifiable risk factor of which the status decreases with age [16] – has been associated with the course of depressive symptoms. In a previous study of the Longitudinal Ageing Study Amsterdam (LASA), an association between low vitamin D status and an increase in depressive symptoms among women was observed [17]. Two recent meta-analyses found a similar association in both men and women [18,19]. Genetic factors, such as single nucleotide polymorphisms (SNPs) in vitamin D-related genes may influence the response to, and therefore the requirement of, vitamin D [20]. The vitamin D receptor (VDR) gene encodes the VDR, the transporter of active vitamin D (1,25(OH)2D), and thereby influences the availability of 1,25(OH)2D [21].

Results

The study sample consisted of 922 participants who were on average 75.6 (SD 6.6) y old, were 51% women, and had a median CES-D score of 13.6 (IQR: 11.6–۱۵٫۶). The prevalence of depressive symptoms [40] and antidepressant use at baseline was 24% and 2.5%, respectively. Mean serum 25(OH)D3 concentration was 46.6 nmol/L (SD 17.7), mean BMI was 26.6 kg/m2 (SD 5.2), and median physical activity in minutes per week was 128.6 (IQR: 67.7–۱۸۹٫۷). The basic characteristics are summarized in Table 1 [35,36,40]. Compared with nonresponders, participants in the study sample were older and more often women, never smokers, moderate-to-excessive alcohol users, and more often had 2 or more comorbidities (all P < 0.05; Supplemental Table 2). The minor allele frequency was smallest for 282 W (n = 63/919; 3%), and no SNPs departed from Hardy–Weinberg equilibrium (Supplemental Table 1).