https://orcid.org/0000-0002-2486-1900
Abstract
Single-nucleotide polymorphisms (SNPs) in the vitamin D receptor (VDR) contribute to inadequate metabolic profiles in individuals with type 2 diabetes mellitus (T2DM).
We sought to elucidate the relationship among SNPs in the VDR and markers for glycemic control, lipid profile, and inflammation in individuals with T2DM.
We performed a systematic search in the MEDLINE (via PubMed), EMBASE, and SCOPUS databases in July 2021 and updated the search in October 2023.
6 observational studies were selected from the databases, and 1 study was included after checking the reference list. Two authors independently completed the selection and data extraction of studies and population characteristics, the prevalence of SNPs in the VDR, genotyping methods, and laboratory findings, and performed summary statistics of the results.
The meta-analyses were performed on 5 studies including 1198 adults with T2DM. The duration of the diabetes diagnosis ranged from 5.0 to 14.7 years. A random-effects model was used to pool the results using a 2-tailed (P < .05). Effect sizes were reported as standardized mean differences (SMDs) and 95% confidence intervals (CIs). Four SNPs in the VDR were identified (Fokl, BsmI, Taql, and Apal) by using polymerase chain reaction–restriction fragment length polymorphism (PCR-RFLP). The Fokl SNP was identified in 5 studies and associated with a higher percentage of glycated hemoglobin (HbA1c%) (SMD, 0.41 [95% CI, 0.15-0.67]). The Bsml in 4 studies was associated with higher triacylglycerol (SMD, 0.21 [95% CI, 0.03-0.38]). The Taql SNP was identified in 2 studies and did not show any associations, and the Apal SNP was identified in only 1 study and was not analysed in the meta-analysis.
Although the studies identified 4 SNPs in the VDR, the results of the meta-analysis allowed us to infer only the association of the SNPs Fokl and Bsml with increased %HbA1c and triacylglycerol levels, respectively, in individuals with T2DM.
PROSPERO registration number CRD42021268152.
INTRODUCTION
Type 2 diabetes mellitus (T2DM) is characterized by insulin resistance, chronic hyperglycemia, and, in some cases, beta cell exhaustion and failure of insulin synthesis and secretion.1 Genetic and environmental factors are associated with an increased risk of T2DM, which is characterized by a complex and multifactorial etiology.2 Obesity is a principal risk factor,3 because adipose tissue dysfunction results in low-grade inflammation and insulin resistance.4 Furthermore, individuals with T2DM commonly present with diabetic dyslipidemia, characterized by an increase in triacylglycerol (TAG), low-density lipoprotein cholesterol (LDL-C), and very low-density lipoprotein cholesterol (VLDL-C) and a decrease in high-density lipoprotein cholesterol (HDL-C).5 Furthermore, vitamin D deficiency and insufficiency are associated with inadequate glycemic control and complications of T2DM,6 and genetic alterations in vitamin D receptors (VDRs) can contribute to a higher risk of developing the disease.7,8
An active form of vitamin D, 1,25 dihydroxyvitamin D (1,25(OH)2D3), binds to VDRs and retinoic acid X receptor (RXR) to form an RXR–VDR heterodimer that recognizes vitamin D response elements (VDRE) and modulates up to 3% of the human genome.9 Furthermore, VDRs have an extensive promoter region capable of generating multiple tissue-specific transcripts. The presence of the promoter region explains the importance of studies on single-nucleotide polymorphisms (SNPs) in the VDR,9 which are located in the 12q13 region and comprise 11 exons.10,11 The SNPs in the VDR gene are associated with poor metabolic control in individuals with T2DM.7,12,13
The SNPs Fokl (rs2228570), Bsml (rs1544410), Taql (rs731236), and Apal (rs7975232) in VDR have been the most studied, especially in Caucasians, Asians, and Africans.13,14 However, the relationship between these SNPs and metabolic control in individuals with T2DM has not been extensively explored, but noteworthy associations have been observed, such as Bsml increasing the risk of metabolic syndrome in women,8 and Fokl and Apal elevating the risk of insulin resistance and the development of T2DM in adults.11,14–16 Even when Taql and Apal are in disequilibrium, they have been associated with an increased risk of T2DM development in adults.15,17,18 A systematic review revealed that the presence of the SNPs Fokl and Bsml in VDR increased the risk of developing diabetic retinopathy in individuals with T2DM, although the SNPs Apal and Taql were not associated with this risk.12
These relationships between SNPs in the VDR and an increased risk of T2DM are widely discussed in the literature; however, it is important to examine the evidence of the influence of these SNPs on metabolic control (glycemic control, lipid profile, and inflammatory markers) in individuals with T2DM.13,19 Thus, in this study we performed a systematic review and meta-analysis of the existing literature to examine the relationship between the presence of SNPs in the VDR and glycemic control, lipid profiles, and inflammatory markers in adults and elderly individuals with T2DM.
METHODS
The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (registration number: CRD42021268152) and followed the recommendations of the Meta-analysis of Observational Studies in Epidemiology (MOOSE) checklist20 (Appendix S1).
Search Strategy
A systematic search was performed in July 2021 and updated in October 2023 in the MEDLINE (via PubMed), EMBASE, and SCOPUS databases. A grey-literature search was conducted using Google Scholar. We did not restrict the year of publication. The search was limited to observational studies published in full text without language restrictions. Studies published in non-English languages were translated using professional translation services when necessary. The search strategy was defined following the research question according to the population, exposure, comparison, outcomes, and study design (PECOS). The PECOS elements are listed in Table 1.
PECOS Anagram: Elements Used to Define the Eligibility Criteria
| Anagram | Item | Description |
|---|---|---|
| P | Population | Adults and elderly with type 2 diabetes |
| E | Exposition | SNP genotypes in the VDR gene |
| C | Comparator | Individuals with type 2 diabetes and the wild-type VDR gene |
| O | Outcomes | Glycemic control, lipid profile and inflammatory markers (HbA1c %, fasting glucose, HOMA-IR, TC, LDL-C, HDL-C, TAG, IL-6, and TNF-alpha) |
| Anagram | Item | Description |
|---|---|---|
| P | Population | Adults and elderly with type 2 diabetes |
| E | Exposition | SNP genotypes in the VDR gene |
| C | Comparator | Individuals with type 2 diabetes and the wild-type VDR gene |
| O | Outcomes | Glycemic control, lipid profile and inflammatory markers (HbA1c %, fasting glucose, HOMA-IR, TC, LDL-C, HDL-C, TAG, IL-6, and TNF-alpha) |
Abbreviations: HbA1c, glycated hemoglobin percentage; HDL-C, high-density lipoprotein cholesterol; HOMA-OR, homeostatic model assessment for insulin resistance; IL-6, interleukin-6; LDL-C, low-density lipoprotein–cholesterol; TAG, triacylglycerol; TC, total cholesterol; TNF-alpha, tumoral nuclear factor alpha.
PECOS Anagram: Elements Used to Define the Eligibility Criteria
| Anagram | Item | Description |
|---|---|---|
| P | Population | Adults and elderly with type 2 diabetes |
| E | Exposition | SNP genotypes in the VDR gene |
| C | Comparator | Individuals with type 2 diabetes and the wild-type VDR gene |
| O | Outcomes | Glycemic control, lipid profile and inflammatory markers (HbA1c %, fasting glucose, HOMA-IR, TC, LDL-C, HDL-C, TAG, IL-6, and TNF-alpha) |
| Anagram | Item | Description |
|---|---|---|
| P | Population | Adults and elderly with type 2 diabetes |
| E | Exposition | SNP genotypes in the VDR gene |
| C | Comparator | Individuals with type 2 diabetes and the wild-type VDR gene |
| O | Outcomes | Glycemic control, lipid profile and inflammatory markers (HbA1c %, fasting glucose, HOMA-IR, TC, LDL-C, HDL-C, TAG, IL-6, and TNF-alpha) |
Abbreviations: HbA1c, glycated hemoglobin percentage; HDL-C, high-density lipoprotein cholesterol; HOMA-OR, homeostatic model assessment for insulin resistance; IL-6, interleukin-6; LDL-C, low-density lipoprotein–cholesterol; TAG, triacylglycerol; TC, total cholesterol; TNF-alpha, tumoral nuclear factor alpha.
The following key search terms were used in the search strategy: “Vitamin D receptor”; “Receptors, calcitriol”; “Polymorphisms”; “Glycemic Control”; “Metabolic Syndrome”; “Lipid”; “Lipid Metabolism”; “Cytokines”; “Inflammation Mediators”; “Inflammation”; “Diabetes Mellitus, Type 2”. The reference lists of eligible studies and reviews were scanned to identify additional studies for inclusion. The search strategy for each database is described in detail in Appendix S2.
Eligibility Criteria
Observational studies developed with individuals with T2DM who presented values of glycemic, lipid, and inflammatory markers according to SNP genotypes in the VDR gene (homozygous and heterozygous vs wild type) were included.
We excluded clinical trials, studies involving animals, case reports, case series, in vitro studies, conference proceedings, scientific meeting abstracts, editorials, letters to the editor that did not provide original data, and studies published incompletely or in abstract form. Studies that included pregnant women or pediatric patients were excluded. The list of studies identified in the database search was checked, and duplicates were manually excluded using the spreadsheet editor Microsoft Excel.
Study Selection and Data Extraction
After checking the list of studies identified in the data search and excluding duplicates, the selection of studies was performed independently by 2 evaluators (R.K.F.S. and R.O.P.) in 2 steps: reading the title and abstract, and reading the full text. Disagreements between the evaluators were resolved by a third evaluator (A.M.O.S.). Microsoft Excel was used for the study selection and data extraction.
To determine the level of concordance between evaluators in the selection and data extraction steps, Cohen’s kappa coefficient was used, considering the range 0-1, being < 0 = no agreement; 0-0.20 = poor agreement; 0.21-0.40 = fair agreement; 0.41-0.60 = moderate agreement; 0.61-0.80 = substantial agreement, and 0.81-1 = almost perfect agreement.21
Strengthening the Reporting of Genetic Association Studies (STREGA),22 an extension of the STROBE Statement, was used to guide the items to be extracted, considering what should be presented in observational studies evaluating the association between diseases and genes. Data were extracted independently by 2 evaluators (R.K.F.S. and R.O.P.).
Thus, we extracted study characteristics (aim and study design), population (country, age, sex), the prevalence of SNPs (Fokl, Bsml, Taql, and Apal) in the VDR, genotyping errors, Hardy–Weinberg equilibrium, replication, selection of participants, rationale for the choice of genes and variants, anthropometric data, and the following laboratory findings: hemoglobin A1c percentage (%HbA1c), fasting glucose, homeostatic model assessment for insulin resistance (HOMA-IR), total cholesterol (TC), LDL-C, HDL-C, triacylglycerol (TAG), interleukin 6 (IL-6), and tumor necrosis factor-alpha (TNF-α).
Methodological Quality Evaluation
Methodological quality was evaluated using the Newcastle–Ottawa Quality Assessment Scale for Case-Control Studies23 and assessments were conducted independently by 2 evaluators (R.K.F.S. and R.O.P.).
Studies were evaluated for participant selection, comparability between cases and controls, and exposure. The following items were observed during participant selection: (1) case-adequate definition, (2) case representative, (3) control selection, and (4) control definition. For the comparability item, a statistically adjusted analysis for age or age difference between the case and control groups was considered. The exposure items were as follows: (1) gold standard methodology for obtaining data, (2) same methods used for both groups, and (3) nonresponse rate.
The study was considered to be of high quality when 4 stars were obtained in the selection, 1 star in comparability, and three stars in the exposure domain.
Statistical Analysis
STATA software (version 14) was used to conduct the meta-analysis. Effect sizes were reported as the standardised mean differences (SMD) with a 95% confidence interval (CI). Additionally, 0.2, 0.5, and 0.8 were considered small, medium, and large effect sizes,24 respectively. A random-effects model was used to pool the results using a 2-tailed test with a significance level of 0.05. Means and standard deviations were extracted from the studies, and the mean differences (MD) were presented. Data presented as CIs were used to calculate the variance and obtain the standard deviation (SD).25
Heterogeneity was investigated using the Cochran Q test with a cut-off of 10% for significance26 and quantified using the I2 index [100% × (Q-df)/Q]27 as follows: 0%, non-heterogeneity between studies; <50%, low heterogeneity; 50%-75%, moderate heterogeneity; >75%, high heterogeneity.28 Forest plots were used to graphically present the significant results of the pooled estimates.
RESULTS
Overall, 1946 reports were identified in the database search, and 1087 duplicates were excluded. After reading the titles and abstracts, 14 studies were reading in full, and 5 were selected for inclusion. One additional study was included after review of the reference lists of the 5 studies already included.29 Thus, 6 case-control studies were included in this systematic review, and the genotype data of individuals with T2DM (group cases) were evaluated.18,29–33 A PRISMA flow diagram34 showing the search and selection strategies is shown in Figure 1.
Flow Diagram of the Search and Selection Strategy of the Studies Included in the Systematic Review
Cohen’s kappa coefficient showed moderate agreement between evaluators in the title and abstract reading phases (Cohen’s kappa = 0.59) and substantial agreement in the full reading phase (Cohen’s kappa = 0.69).
Study Characteristics
In total, 1198 individuals with T2DM living in Korea,18 Iran,31 Chile,30 the United Arab Emirates,33 Egypt,32 and India29 were evaluated in the 6 studies included in this systematic review. Five studies grouped the population according to sex: 53.09% (n = 583) women and 46.91% (n = 515) men.18,30–33 One study did not report the sex of participants.29
The mean age varied from 47.39 (± 6.01) to 62.60 (± 10.60) years, and the diabetes diagnostic time of 5.00 (± 5.70) to 14.70 (± 7.50) years.18,29,31 None of the included studies provided information on the racial demographics of the population.
Glycemic control was evaluated based on fasting glucose concentrations18,29–32 and %HbA1c18,31–33; lipid profile according to TC,18,29,30,32,33 TAG,18,29,30,32,33LDL-C,18,29,32,33 and HDL-C18,29,30,32,33 concentrations; and insulin resistance based on HOMA-IR18,30–32 values. The inflammatory profile was evaluated in 1 study using TNF-α and IL-6 concentrations.32
The main characteristics of the studies and population are shown in Table 2.
General Characteristics of the Population Evaluated in the Studies Included in this Systematic Review
| Reference: Author (year) | Country | Type 2 diabetes, No. | Gender F/M, No. | Age, y | Time of diagnosis, y | BMI, kg/m² | Fasting glucose, mg/dL | HbA1c, % | TC, mg/dL | LDL-C, mg/dL | HDL-C, mg/dL | TAG, mg/dL | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||||
| Nam et al. (2021)18 | Korea | 506 | 240/266 | 62.60 | 10.60 | 14.70 | 7.50 | 25.10 | 3.50 | 145.10 | 55.40 | 7.60 | 1.40 | 170.30 | 33.70 | 96.50 | 28.30 | 45.50 | 11.90 | 158.40 | 100.80 |
| Safar et al. (2017)33 | United Arab Emirates | 264 | 130/134 | 60.50 | 11.59 | NP | 32.12 | 5.93 | NE | 7.48 | 1.36 | 159.70 | 61.48a | 84.30 | 44.08a | 47.95 | 23.59a | 160.32 | 186.90a | ||
| Angel et al. (2016)30 | Chile | 160 | 102/58 | 61.70 | 11.50 | NP | 31.00 | 5.60 | 143.10 | 88.10 | NP | 195.70 | 46.10 | NP | 33.60 | 10.50 | 147.20 | 65.20 | |||
| Mackawy and Badawi, (2014)32 | Egypt | 63 | 27/36 | 47.39 | 6.01 | NP | 30.04 | 2.94 | 138.86 | 7.91 | 11.37 | 2.46 | 288.55 | 69.04 | 228.22 | 22.03 | 43.00 | 4.43 | 177.80 | 65.20 | |
| Hossein-nezhad (2009)31 | Iran | 105 | 84/21 | 55.00 | 10.00 | 5.58 | 4.25 | 28.99 | 4.28 | NE | 7.40 | 1.80 | NP | NP | NP | NP | |||||
| Bid et al. (2009)29 | India | 100 | NP | 49.32 | 10.97 | 5.00 | 5.70 | 24.26 | 4.30 | 174.30 | 79.44 | NP | 225.13 | 33.10 | 160.80 | 30.41 | 42.60 | 3.50 | 115.00 | 14.10 | |
| Reference: Author (year) | Country | Type 2 diabetes, No. | Gender F/M, No. | Age, y | Time of diagnosis, y | BMI, kg/m² | Fasting glucose, mg/dL | HbA1c, % | TC, mg/dL | LDL-C, mg/dL | HDL-C, mg/dL | TAG, mg/dL | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||||
| Nam et al. (2021)18 | Korea | 506 | 240/266 | 62.60 | 10.60 | 14.70 | 7.50 | 25.10 | 3.50 | 145.10 | 55.40 | 7.60 | 1.40 | 170.30 | 33.70 | 96.50 | 28.30 | 45.50 | 11.90 | 158.40 | 100.80 |
| Safar et al. (2017)33 | United Arab Emirates | 264 | 130/134 | 60.50 | 11.59 | NP | 32.12 | 5.93 | NE | 7.48 | 1.36 | 159.70 | 61.48a | 84.30 | 44.08a | 47.95 | 23.59a | 160.32 | 186.90a | ||
| Angel et al. (2016)30 | Chile | 160 | 102/58 | 61.70 | 11.50 | NP | 31.00 | 5.60 | 143.10 | 88.10 | NP | 195.70 | 46.10 | NP | 33.60 | 10.50 | 147.20 | 65.20 | |||
| Mackawy and Badawi, (2014)32 | Egypt | 63 | 27/36 | 47.39 | 6.01 | NP | 30.04 | 2.94 | 138.86 | 7.91 | 11.37 | 2.46 | 288.55 | 69.04 | 228.22 | 22.03 | 43.00 | 4.43 | 177.80 | 65.20 | |
| Hossein-nezhad (2009)31 | Iran | 105 | 84/21 | 55.00 | 10.00 | 5.58 | 4.25 | 28.99 | 4.28 | NE | 7.40 | 1.80 | NP | NP | NP | NP | |||||
| Bid et al. (2009)29 | India | 100 | NP | 49.32 | 10.97 | 5.00 | 5.70 | 24.26 | 4.30 | 174.30 | 79.44 | NP | 225.13 | 33.10 | 160.80 | 30.41 | 42.60 | 3.50 | 115.00 | 14.10 | |
Abbreviations: BMI, body mass index; F, female; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein–cholesterol; LDL-C, low-density lipoprotein–cholesterol; M, male; NE, not evaluated; NP, not presented; TAG, triacylglycerol; TC, total cholesterol.
Values obtained in mmol/L and converted to mg/dL using an online calculator: TC, LDL-C, and HDL-C values multiply by 38.67; TAG multiply by 88.57.
General Characteristics of the Population Evaluated in the Studies Included in this Systematic Review
| Reference: Author (year) | Country | Type 2 diabetes, No. | Gender F/M, No. | Age, y | Time of diagnosis, y | BMI, kg/m² | Fasting glucose, mg/dL | HbA1c, % | TC, mg/dL | LDL-C, mg/dL | HDL-C, mg/dL | TAG, mg/dL | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||||
| Nam et al. (2021)18 | Korea | 506 | 240/266 | 62.60 | 10.60 | 14.70 | 7.50 | 25.10 | 3.50 | 145.10 | 55.40 | 7.60 | 1.40 | 170.30 | 33.70 | 96.50 | 28.30 | 45.50 | 11.90 | 158.40 | 100.80 |
| Safar et al. (2017)33 | United Arab Emirates | 264 | 130/134 | 60.50 | 11.59 | NP | 32.12 | 5.93 | NE | 7.48 | 1.36 | 159.70 | 61.48a | 84.30 | 44.08a | 47.95 | 23.59a | 160.32 | 186.90a | ||
| Angel et al. (2016)30 | Chile | 160 | 102/58 | 61.70 | 11.50 | NP | 31.00 | 5.60 | 143.10 | 88.10 | NP | 195.70 | 46.10 | NP | 33.60 | 10.50 | 147.20 | 65.20 | |||
| Mackawy and Badawi, (2014)32 | Egypt | 63 | 27/36 | 47.39 | 6.01 | NP | 30.04 | 2.94 | 138.86 | 7.91 | 11.37 | 2.46 | 288.55 | 69.04 | 228.22 | 22.03 | 43.00 | 4.43 | 177.80 | 65.20 | |
| Hossein-nezhad (2009)31 | Iran | 105 | 84/21 | 55.00 | 10.00 | 5.58 | 4.25 | 28.99 | 4.28 | NE | 7.40 | 1.80 | NP | NP | NP | NP | |||||
| Bid et al. (2009)29 | India | 100 | NP | 49.32 | 10.97 | 5.00 | 5.70 | 24.26 | 4.30 | 174.30 | 79.44 | NP | 225.13 | 33.10 | 160.80 | 30.41 | 42.60 | 3.50 | 115.00 | 14.10 | |
| Reference: Author (year) | Country | Type 2 diabetes, No. | Gender F/M, No. | Age, y | Time of diagnosis, y | BMI, kg/m² | Fasting glucose, mg/dL | HbA1c, % | TC, mg/dL | LDL-C, mg/dL | HDL-C, mg/dL | TAG, mg/dL | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||||
| Nam et al. (2021)18 | Korea | 506 | 240/266 | 62.60 | 10.60 | 14.70 | 7.50 | 25.10 | 3.50 | 145.10 | 55.40 | 7.60 | 1.40 | 170.30 | 33.70 | 96.50 | 28.30 | 45.50 | 11.90 | 158.40 | 100.80 |
| Safar et al. (2017)33 | United Arab Emirates | 264 | 130/134 | 60.50 | 11.59 | NP | 32.12 | 5.93 | NE | 7.48 | 1.36 | 159.70 | 61.48a | 84.30 | 44.08a | 47.95 | 23.59a | 160.32 | 186.90a | ||
| Angel et al. (2016)30 | Chile | 160 | 102/58 | 61.70 | 11.50 | NP | 31.00 | 5.60 | 143.10 | 88.10 | NP | 195.70 | 46.10 | NP | 33.60 | 10.50 | 147.20 | 65.20 | |||
| Mackawy and Badawi, (2014)32 | Egypt | 63 | 27/36 | 47.39 | 6.01 | NP | 30.04 | 2.94 | 138.86 | 7.91 | 11.37 | 2.46 | 288.55 | 69.04 | 228.22 | 22.03 | 43.00 | 4.43 | 177.80 | 65.20 | |
| Hossein-nezhad (2009)31 | Iran | 105 | 84/21 | 55.00 | 10.00 | 5.58 | 4.25 | 28.99 | 4.28 | NE | 7.40 | 1.80 | NP | NP | NP | NP | |||||
| Bid et al. (2009)29 | India | 100 | NP | 49.32 | 10.97 | 5.00 | 5.70 | 24.26 | 4.30 | 174.30 | 79.44 | NP | 225.13 | 33.10 | 160.80 | 30.41 | 42.60 | 3.50 | 115.00 | 14.10 | |
Abbreviations: BMI, body mass index; F, female; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein–cholesterol; LDL-C, low-density lipoprotein–cholesterol; M, male; NE, not evaluated; NP, not presented; TAG, triacylglycerol; TC, total cholesterol.
Values obtained in mmol/L and converted to mg/dL using an online calculator: TC, LDL-C, and HDL-C values multiply by 38.67; TAG multiply by 88.57.
Genotyping was performed using polymerase chain reaction-restriction fragment-length polymorphisms (PCR-RFLP). The identified SNPs were Fokl (rs2228570) (ff) (five studies29–33), Bsml (rs1544410) (BB) (4 studies18,29,32,33), Taql (rs731236) (TT) (2 studies29,33), and Apal (rs7975232) (AA) (evaluated in a single study).18 In all studies, the SNPs were in Hardy-Weinberg equilibrium. The frequencies of the SNPs evaluated in each study are shown in Table S1.
In individual analyses, the SNP Fokl was associated with higher values of glycemic control markers, lipid profiles, and IL-6, whereas the SNPs Bsml and Taql were associated with inadequate lipid profiles. The SNP Apal was not associated with an increase in the evaluated markers. The main findings of these studies are summarised in Table 3.
Main Findings on the Relationship Between Vitamin D Receptor Gene Polymorphisms and Metabolic Markers
| Genotype | Author (year) | Main finding |
|---|---|---|
| Fokl rs 2228570 | ||
| ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c and HOMA-IR |
| Mackway and Badawi, (2014)32 | Associated with higher TAG, LDL-C, IL-6, and lower HDL-C | |
| Ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c |
| Angel et al. (2016)30 | Associated with higher TAG and HOMA-IR | |
| Mackway and Badawi, (2014)32 | Associated with higher TC | |
| Bsml rs1544410 | ||
| Bb | Safar et al. (2017)33 | Associated with higher LDL-C |
| Taql rs731236 | ||
| TT | Safar et al. (2017)33 | Associated with higher TC and LDL-C |
| Tt | Safar et al. (2017)33 | Associated with TC and LDL-C |
| Genotype | Author (year) | Main finding |
|---|---|---|
| Fokl rs 2228570 | ||
| ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c and HOMA-IR |
| Mackway and Badawi, (2014)32 | Associated with higher TAG, LDL-C, IL-6, and lower HDL-C | |
| Ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c |
| Angel et al. (2016)30 | Associated with higher TAG and HOMA-IR | |
| Mackway and Badawi, (2014)32 | Associated with higher TC | |
| Bsml rs1544410 | ||
| Bb | Safar et al. (2017)33 | Associated with higher LDL-C |
| Taql rs731236 | ||
| TT | Safar et al. (2017)33 | Associated with higher TC and LDL-C |
| Tt | Safar et al. (2017)33 | Associated with TC and LDL-C |
Abbreviations: ff and TT, homozygote polymorphic; Ff, Bb, and Tt, heterozygote polymorphic; HDL, high-density lipoprotein; HOMA-IR, homeostatic model asessment for insulin resistance; IL-6, interleukin 6; LDL-C, low-density lipoprotein cholesterol; TAG, triacylglycerol; TC, total cholesterol; %HbA1c, hemoglobin A1c percentage.
Main Findings on the Relationship Between Vitamin D Receptor Gene Polymorphisms and Metabolic Markers
| Genotype | Author (year) | Main finding |
|---|---|---|
| Fokl rs 2228570 | ||
| ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c and HOMA-IR |
| Mackway and Badawi, (2014)32 | Associated with higher TAG, LDL-C, IL-6, and lower HDL-C | |
| Ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c |
| Angel et al. (2016)30 | Associated with higher TAG and HOMA-IR | |
| Mackway and Badawi, (2014)32 | Associated with higher TC | |
| Bsml rs1544410 | ||
| Bb | Safar et al. (2017)33 | Associated with higher LDL-C |
| Taql rs731236 | ||
| TT | Safar et al. (2017)33 | Associated with higher TC and LDL-C |
| Tt | Safar et al. (2017)33 | Associated with TC and LDL-C |
| Genotype | Author (year) | Main finding |
|---|---|---|
| Fokl rs 2228570 | ||
| ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c and HOMA-IR |
| Mackway and Badawi, (2014)32 | Associated with higher TAG, LDL-C, IL-6, and lower HDL-C | |
| Ff | Hossein-nezhad et al. (2009)31 | Associated with higher %HbA1c |
| Angel et al. (2016)30 | Associated with higher TAG and HOMA-IR | |
| Mackway and Badawi, (2014)32 | Associated with higher TC | |
| Bsml rs1544410 | ||
| Bb | Safar et al. (2017)33 | Associated with higher LDL-C |
| Taql rs731236 | ||
| TT | Safar et al. (2017)33 | Associated with higher TC and LDL-C |
| Tt | Safar et al. (2017)33 | Associated with TC and LDL-C |
Abbreviations: ff and TT, homozygote polymorphic; Ff, Bb, and Tt, heterozygote polymorphic; HDL, high-density lipoprotein; HOMA-IR, homeostatic model asessment for insulin resistance; IL-6, interleukin 6; LDL-C, low-density lipoprotein cholesterol; TAG, triacylglycerol; TC, total cholesterol; %HbA1c, hemoglobin A1c percentage.
Meta-Analysis
To better understand the results, we standardised the SNPs nomenclature as follows: Fokl genotypes Ff (cytosine>thymine) and ff, BsmI genotypes Bb (adenine> guanine) and BB, Taql genotypes Tt (thymine > cytosine) and TT, and Apal genotypes Aa (adenine>cytosine) and AA.
To evaluate the relationship between the genotypes and biomarkers of glucose control and lipid profile, the analysis was performed by grouping the individuals into 2 groups: Homozygous and heterozygous genotypes vs the homozygous wild-type.
The Apal SNP was not included in the meta-analysis because of insufficient data (only 1 study).18 Furthermore, the SNP Taql (rs731236) was not associated with metabolic outcomes.29,33 The meta-analysis results for all the evaluated SNPs are described in the Table S2.
SNP Fokl (rs2228570)
Meta-analysis showed that genotypes ff and Ff were associated with higher %HbA1c values when compared to the genotype wild (FF) (MD: 8.52 ± 1.88% vs 7.65 ± 1.54%, respectively) (SMD = 0.41, 95% CI, 0.15-0.67, P = .002), considering homogeneous studies and small to moderate effect size31–33 (Figure 2). Nonetheless, ff and Ff genotypes were not associated with differences in fasting glucose, HOMA-IR values, or TC, TAG, HDL-C, and LDL-C concentrations (Table S2).
Meta-Analysis of the Relationship Between the SNP Fokl and Glycated Hemoglobin Percentage (%HbA1c) (A) and SNP Bsml and Triacylglycerol Concentrations (B) in Individuals with Type 2 Diabetes.
SNP Bsml (rs1544410)
Genotypes BB and Bb were associated with higher TAG levels when compared to the genotype wild (bb) (MD: 165.68 ± 106.13 mg/dL vs 150.91 ± 155.55 mg/dL, respectively) (SMD = 0.21, 95% CI, 0.03–0.38, P = .023), considering homogeneous studies and small effect size18,32,33 (Figure 2). Genotypes were not associated with differences in fasting glucose levels, HOMA-IR values, %HbA1C, TC, HDL-C, or LDL-C concentrations (Table S2).
Methodological Quality Evaluation
All studies included cases and control groups that were correctly defined and evaluated using gold standard methods. The comparability between groups was satisfactory in three studies.18,30,32 One study did not receive stars for domain comparability because the mean age was significantly different between groups,33 and in 2 studies, a comparability assessment was not performed.29,31 Moreover, the results reported were not affected by the nonresponse rate, which was the same for both groups. Table 4 presents the results of the bias risk assessment of the studies included in the systematic review.
Assessment of Methodological Quality of the Studies Included in this Systematic Review According to the Newcastle–Ottawa Quality Assessment Scale Case-Control Studies
| Author (year) | Selection | Comparability | Exposure |
|---|---|---|---|
| Nam et al.18 |  | | |
| Safar et al. (2017)33 | | | |
| Angel et al. (2016)30 | | | |
| Mackway and Badawi (2014)32 | | | |
| Bid et al. (2009) 31 | | | |
| Hossein-Nezhad et al. (2009)29 | | |
| Author (year) | Selection | Comparability | Exposure |
|---|---|---|---|
| Nam et al.18 | | | |
| Safar et al. (2017)33 | | | |
| Angel et al. (2016)30 | | | |
| Mackway and Badawi (2014)32 | | | |
| Bid et al. (2009) 31 | | | |
| Hossein-Nezhad et al. (2009)29 | | |
Note: According to the NOS for case-control studies, a maximum of 9 stars can be assigned: up to 4 stars for Selection, up to 2 stars for Comparability, and up to 3 stars for Exposure. Higher numbers of stars indicate better methodological quality in each domain.
Assessment of Methodological Quality of the Studies Included in this Systematic Review According to the Newcastle–Ottawa Quality Assessment Scale Case-Control Studies
| Author (year) | Selection | Comparability | Exposure |
|---|---|---|---|
| Nam et al.18 | | | |
| Safar et al. (2017)33 | | | |
| Angel et al. (2016)30 | | | |
| Mackway and Badawi (2014)32 | | | |
| Bid et al. (2009) 31 | | | |
| Hossein-Nezhad et al. (2009)29 | | |
| Author (year) | Selection | Comparability | Exposure |
|---|---|---|---|
| Nam et al.18 | | | |
| Safar et al. (2017)33 | | | |
| Angel et al. (2016)30 | | | |
| Mackway and Badawi (2014)32 | | | |
| Bid et al. (2009) 31 | | | |
| Hossein-Nezhad et al. (2009)29 | | |
Note: According to the NOS for case-control studies, a maximum of 9 stars can be assigned: up to 4 stars for Selection, up to 2 stars for Comparability, and up to 3 stars for Exposure. Higher numbers of stars indicate better methodological quality in each domain.
DISCUSSION
The present meta-analysis suggests that Fokl and Bsml SNPs were associated with higher %HbA1c and TAG concentrations, respectively. Associations between SNPs and other variables were not observed, which could be explained by the small number of studies and size of the studied population. The small effect size in the meta-analysis suggests that environmental factors, such as diet, time of diagnosis of T2DM, medications, and race, may influence the outcomes of noncommunicable diseases, similar to genetic factors. A previous study proposed that environmental factors contribute to 70%–90% of disease risks,35 which could explain our results.
Population homogeneity has a significant effect on the results of genetic studies, as the prevalence of SNPs in the VDR is associated with the race of the individuals. Asians presented a higher prevalence of SNP Fokl and Apal, whereas Caucasians and Africans showed a higher prevalence of SNPs Fokl, Bsml, Apal, and Taql.9 In this systematic review, the SNPs Fokl and Bsml were the most prevalent among the studies. However, the included studies did not describe the races of the evaluated populations.
Although the studies included in this meta-analysis included populations with elevated %HbA1c levels, the Fokl SNP was associated with higher %HbA1c concentrations. Thus, Fokl may be considered an important risk factor for increased HbA1c levels in individuals with T2DM.31
The SNP Fokl is related to reduced synthesis of messenger RNA (mRNA)36 and presents a longer protein with 427 amino acids, but lower activity compared to the wild genotype (424 amino acids).37 Furthermore, the Fokl SNP did not show linkage disequilibrium with the other SNPs in the VDR. Therefore, it can be an independent factor for phenotypic manifestation.37
The lower activity of the VDR protein observed in the presence of Fokl modifies the functional properties of the receptor, decreasing the nuclear response to 1,25(OH)D.38 The presence of VDR in pancreatic beta cells and several insulin-dependent tissues and organs justifies its relationship with glucose metabolism and the inflammatory response.11
Additionally, the lower response of VDR in beta cells decreases their capacity to convert pro-insulin into insulin,19 whereas it is associated with peripheral insulin resistance in musculoskeletal and adipose tissues.39,40
The association between VDR and insulin resistance is explained by the role played by VDR in decreasing the inflammatory response by directly interacting with the inhibitor of nuclear factor kappa-B kinase (IKKβ) protein, inhibiting p65 nuclear translocation, and suppressing RelB transcription, blocking the canonical nuclear factor kappa b (NF-κB) pathway.41 Corroborating the mechanism described, the Fokl SNP was associated with an increase in IL-6 in 1 of the studies included in this systematic review, showing that lower activity of the VDR protein interferes with cytokine synthesis, further contributing to the inflammatory state in T2DM.31 Furthermore, vitamin D deficiency is associated with reduced insulin secretion and insulin resistance in humans.36,42–44 1,25(OH)D linked to VDR contributes to insulin secretion and action through calcium channel modulation in beta cells and modulation of calcium content in muscle cells, which causes dephosphorylation of glucose transporter 4 (GLUT-4).45 The studies included in this review did not show vitamin D levels, which limited extrapolation for a better understanding of the relationship between the SNP Fokl, vitamin D levels, and glycemic control markers.
Our meta-analysis showed that SNP Bsml in the VDR was associated with increased TAG concentrations. Studies evaluating the relationship between the SNP Bsml and poor lipid control in individuals with T2DM are controversial.18,29 However, in adult Caucasian women with vitamin D deficiency or insufficiency, the Bsml SNP was associated with increased TAG concentrations and the Taql SNP was associated with higher TC and LDL-C concentrations.46
Some mechanisms explain the relationship between SNPs in the VDR and lipid metabolism. VDR expression in adipose cells increases intracellular calcium levels and decreases the level of cyclic adenosine monophosphate (cAMP), thereby reducing hormone-sensitive lipase and adipose triglyceride lipase gene expression in rats.47 The VDR regulates the gene expression enzymes of the lipogenesis process such as fatty acid synthase (FASN), fatty acid binding protein (FABP), and peroxisome proliferator activator receptor gamma (PPAR-γ), regulating lipoprotein lipase (LPL) in human subcutaneous preadipocytes.48,49 Figure 3 summarizes the main mechanisms that explain the relationship between the SNPs Fokl and Bsml in VDR and inadequate glycemic control and lipid profile in individuals with T2DM.
Main Mechanisms Explaining the Relationship Between SNPs Fokl and Bsml in VDR and Inadequate Glycemic Control and Lipid Profile in Individuals with Type 2 Diabetes Mellitus. Vitamin D receptors (VDRs) were identified in cells of insulin-dependent tissues. The single-nucleotide polymorphism (SNP) Fokl (rs2228570) modifies the functional properties of the receptor, decreasing the nuclear response of the 1,25 dihydroxyvitamin D (1,25(OH)D), reducing the conversion of the proinsulin into insulin in pancreatic cells and VDR action in the inflammatory response by increasing the pro-inflammatory cytokines as interleukin 6 (IL-6) via interaction with the inhibitor of nuclear factor kappa-B kinase (IKKβ) protein, p65 nuclear translocation, and RelB transcription, and block of the canonical (nuclear factor kappa b). The NF-κB pathway contributes to both mechanisms (decrease of insulin secretion and action) and explain the relationship between the SNP Fokl and higher glycated hemoglobin percentage (%HbA1c) in individuals with type 2 diabetes mellitus. SNP Bsml (rs1544410) modulates lipid metabolism by reducing VDR expression in adipose cells, the level of cyclic adenosine monophosphate (cAMP), and the gene expressions of hormone-sensitive lipase and adipose triglyceride lipase. Additionally, the SNP Bsml modifies the contribution of the VDR in regulation of the gene expression enzymes of the lipogenesis process, such as fatty acid synthase (FASN), fatty acid binding protein (FABP), peroxisome proliferator activator receptor (PPAR)-γ, and lipoprotein lipase (LPL) in human subcutaneous preadipocytes, thus contributing to the increase of triacylglycerol (TAG) in individuals with type 2 diabetes mellitus. ↑ indicates increase; ↓ indicates decrease.
In contrast, no difference was observed between the metabolic control markers in individuals with the SNPs Apal and Taql in the VDR compared with those with the wild genotype. A possible explanation for this finding is the position in intron 8, which allows the SNP Apal to create alternative splicing, including 1 exon in the noncoding RNA.9,50 The localization of Bsml, Taql, and Apal in the VDR contributes to the development of haplotypes with important biological effects. They are localized in a regulated region (3UTR) and result in different translations of mRNA, creating haplotypes Bsml+Taql+Apal and, consequently, the expression of Fokl variant proteins with different functions.9,51
The SNPs Fokl and Bsml and the haplotypes (Taq + ApaI) in the VDR were associated with higher susceptibility to T2DM in the Emirates but not with increased LDL-C concentrations in individuals with T2DM of the same ethnicity.33 Furthermore, the SNP Taql and haplotypes (Taql + Bsml) in the VDR were associated with an increased risk of developing diabetic nephropathy. However, this result was not observed when the SNP Taql was isolated.51
Strengths and Limitations
This meta-analysis was conducted according to MOOSE and registered in the PROSPERO database to ensure comprehensive and transparent information. A systematic search was conducted in an expressive number of the databases, according to PECOS, independently by 2 authors, and robust statistical analysis was used to report the associations observed, ensuring methodological robustness. Moreover, this meta-analysis highlights the discussion about the isolated relationship between SNPs in VDR and inadequate glycemic control and lipid profiles in individuals with T2DM living on different continents. The good methodological quality and homogeneity of the studies included in the meta-analysis also supported these outcomes.
The limitations of the number of studies included may explain the lack of associations between SNPs and the other variables analysed and the small sample size. Furthermore, environmental factors such as diet, sun exposure, 25(OH)D3 concentrations, and medications are important for metabolic control in individuals with T2DM but have not been evaluated in previous studies.
CONCLUSIONS
The SNPs Fokl, Bsml, Taql, and Apal in VDR were identified in individuals with T2DM; however, only the SNPs Fokl and Bsml were associated with increased %HbA1c and TAG concentrations, respectively. Our meta-analysis did not allow us to infer associations between the SNPs Taql and Apal and glycemic control markers, lipid profiles, and inflammatory markers in individuals with T2DM. Thus, SNPs in VDR are individual factors that contribute to inadequate glycemic control and lipid profiles in individuals with T2DM. Observational studies developed in other regions of the world with larger sample sizes and evaluations of the SNPs Taql and Apal in the VDR are necessary to confirm the relationship between the main SNPs identified in individuals with T2DM and metabolic control in this population.
Author Contributions
R.K.F.S. and A.M.O.S. equally contributed significantly to the work's conception, design, drafting and editing of the original draft. R.K.F.S. and R.O.P. contributed significantly to study selection and data collection. P.N.B.L. contributed substantially to data analysis. P.R.M.F. contributed substantially to the data interpretation and analysis. R.K.F.S., L.V.P., P.N.B.L., and P.R.M.F. contributed to critical review and editing of the manuscript. A.M.O.S. and L.V.P. contributed to study supervision. All authors read and approved the final version of the manuscript. All authors share responsibility for ensuring that the manuscript complies with the journal's style requirements and terms of consideration.
Supplementary Material
Supplementary Material is available at Nutrition Reviews online.
Funding
This work was supported by the Coordination for the Improvement of Higher Education Personnel (CAPES/FAPITEC N° 10/2016 - PROMOB Process: 88881.157967/2017–01 and 8887.507820/2020–00) and the São Paulo Research Foundation (FAPESP) (grant number 2019/22934–1).
Conflicts of Interest
None declared.
Data Availability
Data available on reasonable request.
