The association between the CCDC88A gene polymorphism at rs1437396 and alcohol use disorder, with or without major depression disorder
, , e | 26 giu 2023
Article Category: Original article
Pubblicato online: 26 giu 2023
Pagine: 127 - 133
Ricevuto: 01 nov 2022
Accettato: 01 mag 2023
DOI: https://doi.org/10.2478/aiht-2023-74-3690
© 2023 Maria Bonea et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
Alcohol use disorder (AUD) is one of the major causes of morbidity and mortality, affecting more than 95 million individuals globally (1, 2). Apart from psychological and environmental factors, around 45–65 % of AUD cases is owed to genetic factors (3). So far, AUD has been associated with numerous single nucleotide polymorphisms (SNPs) in genes encoding alcohol-metabolising enzymes [alcohol dehydrogenase (ADH) and acetaldehyde dehydrogenase (ALDH)] and most neurotransmitters (such as serotonin, dopamine, and glutamate), each with an effect size of less than 1 % (R2<0.005) (4, 5). The most studied are the associations between AUD and the
One genome-wide association study with 16,087 participants (21) described a new significant association between AUD and the rs1437396 polymorphism (p=1.17 x 10-10), located on the short arm of chromosome 2 (2p) in an intergenic region, 9.5 kb downstream from the transcription stop site of the coiled-coil domain-containing 88A (
The aim of this study was to find the association between the
Participants were recruited consecutively at the Psychiatry Department of the County Clinical Emergency Hospital in Cluj-Napoca, Romania between May 2017 and June 2020. In total, the study included 617 participants (417 men and 200 women), of whom 226 were AUD (53 AUD+MDD) patients (excluded were those who did not want to participate or were diagnosed psychiatric disorders other than AUD and/or MDD) and 391 healthy controls with no history of AUD or other psychiatric disorders. AUD and MDD were diagnosed in a structured clinical interview using the DSM-5 classification (29). The severity of depressive symptoms was scored on the 17-item Hamilton Depression Rating Scale (HAM-D) (30) as mild (10–13 points), mild to moderate (14–17 points), or moderate to severe (>17 points).
All participants filled out a questionnaire with their demographic data, including age, residence (urban/rural), education, employment (employed/unemployed/disability pension/retired), marital status (single/with a partner), and family history of AUD and signed informed consent before the study began. The study was approved by the Ethics Committee of the Iuliu Hatieganu University of Medicine and Pharmacy (approval No. 167/07.04.2017) and complied with the principles of the Helsinki Declaration.
Blood (2 mL) was taken from all patients by venipuncture into vacutainers coated with ethylene-diamine-tetra-acetic acid (EDTA) to prevent coagulation. The samples were stored at +4 °C for a maximum of two days until processing. DNA was extracted according to the protocol of the commercial kit producer Wizzard® Genomic DNA Purification Kit (Promega, Madison, WA, USA). The genotypes of the rs1437396 polymorphism were determined using the TaqMan SNP Genotyping assay run on a QuantStudio 3 real-time polymerase chain reaction (PCR) machine (both from Applied Biosystems, Thermo Fisher, Waltham, MA, USA). The PCR primers were as follows: rs1437396 forward, 5’-AAA AGA ATG ACA TTT AGT TAT GTA-3’ and rs1437396 reverse, 3’-AAT GAA TTA TAT GAG CTT TTT TTG-5’.
The studied alleles were T (the risk allele) and C, forming the TT, TC, and CC genotypes.
Data were processed using the IBM SPSS Statistics 25.0 (IBM, Armonk, NY, USA), and we relied on the chi-squared test for the Hardy-Weinberg equilibrium (HWE). The difference in allelic and genotype distributions between patients and controls was assessed with Fisher’s exact test, assuming three genetic models (co-dominant, dominant, and recessive). Binary univariate logistic regression analysis was used to estimate the additional risk of comorbid depression. The statistical power was assessed with CaTS Power Calculator for Genetic Studies (31, 32) and significance set to p<0.05.
The median age of the AUD group was 51 (IQR 47) and of controls 45 (IQR 39). Table 1 shows the demographic and clinical information about the AUD+MDD and AUD-only patients. The allele distribution was normal and in line with the Hardy-Weinberg equilibrium for both the AUD group (χ2=0.1, p=0.75) and controls (χ2=1.57, p=0.78), indicating that the sample is representative of the population (p>0.05).
Demographic and clinical characteristics of alcohol use disorder (AUD) patients with or without major depressive disorder (AUD+MDD)
| Demographic data | AUD+MDD N=53 (%) | AUD only N=173 (%) | p-value | |
|---|---|---|---|---|
| Age | 49 (35) | 52 (48) | 0.178a | |
| Gender | Female | 7 (13.2) | 27 (13.3) | 1c |
| Male | 46 (86.8) | 150 (86.7) | ||
| Living environment | Urban | 35 (66) | 99 (57.2) | 0.261c |
| Rural | 18 (34) | 74 (42.8) | ||
| Education | College | 8 (15) | 5 (2.9) | |
| High school | 20 (37.4) | 53 (30.6) | ||
| Vocational school | 15 (28.3) | 78 (45.1) | ||
| Elementary school | 10 (18.3) | 37 (21.4) | ||
| Marital status | With a partner | 26 (49) | 101 (58.4) | 0.261c |
| Single | 27 (51) | 72 (41.6) | ||
| Family history of AUD | Yes | 34 (64.1) | 114 (65.9) | 0.867c |
| No | 19 (35.9) | 59 (34.1) | ||
| Occupation | Employed | 24 (45.3) | 58 (33.5) | 0.052b |
| Unemployed | 9 (17) | 52 (30.1) | ||
| Disability pension | 14 (26.4) | 29 (16.8) | ||
| Retirement due to age | 6 (11.3) | 34 (19.6) |
Age (continuous variable) is presented as median with interquartile range in parentheses. Categorical variables are reported as number of participants with percentage in parenthesis.
Mann-Whitney
chi-squared test,
Fisher’s exact test. Statistically significant p-values are bolded and marked with asterisk. AUD – alcohol use disorder; MDD – major depressive disorder
Table 2 shows the distribution of the SNP rs1437396 genotypes in the AUD group and controls. The T allele of polymorphism rs1437396 turned out to be significantly associated with a higher risk of AUD in dominant or codominant models of transmission. After applying binary logistic regression, the dominant model explains 1.1 % of the risk of developing AUD (Nagelkerke
Distribution of genotype and allele frequency of the rs1437396 polymorphism in patients with alcohol use disorder (AUD) and healthy controls
| Model | AUD N=226 (%) | Controls N=391 (%) | OR (95 % CI) | p-value | |
|---|---|---|---|---|---|
| Allelic | T vs C | 120 (26.5) | 167 (21.4) | 1.33 | |
| 332 (73.5) | 615 (78.6) | (1.01–1.14) | |||
| Co-dominant | TT vs CC | 15 (6.6) | 22 (5.63) | 1.38 | |
| 121 (53.5) | 246 (62.9) | (0.7–2.7) | 0.36 | ||
| CT vs CC | 90 (39.8) | 123 (31.5) | 1.48 | ||
| 121 (53.5) | 246 (62.9) | (1.05–2.1) | |||
| Dominant | TT+CT vs CC | 105 (46.4) | 145 (37) | 1.47 | |
| 121 (53.6) | 246 (63) | (1.05–2.05) | |||
| Recessive | TT vs CC+CT | 15 (6.6) | 22 (5.6) | 1.19 | |
| 211 (63.4) | 369 (94.4) | (0.6–2.34) | 0.6 |
p-values were determined by Fisher’s exact test. Statistically significant p-values are bolded and marked with the asterisk
Table 3 shows the significantly higher frequency of the T allele for AUD+MDD patients than for those without MDD. This association is confirmed by log regression analysis, used also to evaluate other variables potentially contributing to MDD (Table 4). Of them, only high (college) education showed significant association with AUD. The regression pattern explains 1.14 % of cases of the dual diagnosis (Nagelkerke R2=0.114) and can predict the correct categorization in 77.9 % of cases. The Hosmer-Lemeshow test confirms that the regression model is adequate for predicting depression comorbid with AUD (p=0.819).
Distribution of genotype and allele frequency of the rs1437396 polymorphism in patients with alcohol use disorder and major depressive disorder (AUD+MDD) vs the AUD-only group
| Model | AUD+MDD N=53 (%) | AUD N=173 (%) | OR (95 % CI) | p-value | |
|---|---|---|---|---|---|
| Allelic | T vs C | 38 (35.85) | 82 (23.7) | 1.8 | |
| 68 (64.15) | 264 (76.3) | (1.12–2.87) | |||
| Co-dominant | TT vs CC | 6 (11.32) | 9 (5.2) | 3.17 | 0.077 |
| 21 (39.62) | 100 (57.8) | (1.02–9.88) | |||
| CT vs CC | 26 (49.05) | 64 (37) | 1.93 | 0.065 | |
| 21 (39.62) | 100 (57.8) | (1–3.72) | |||
| Dominant | TT+CT vs CC | 32 (60.38) | 73 (42.2) | 2.08 | |
| 21 (39.62) | 100 (57.8) | (1.11–3.91) | |||
| Recessive | TT vs CC+CT | 6 (11.32) | 9 (5.2) | 2.32 | 0.124 |
| 47 (88.68) | 164 (94.8) | (0.78–6.67) |
p-values were determined by Fisher’s exact test. Statistically significant p-values are bolded and marked with asterisk
Logistic regression for the association between the rs1437396 polymorphism and alcohol use disorder and major depressive disorder (AUD+MDD) comorbidity (N=53)
| Variables | B | SE | df | OR (CI 95 %) | p-value | |
|---|---|---|---|---|---|---|
| rs1437396 (T allele) | 0.779 | 0.338 | 5.326 | 1 | 2.17 (1.12–4.22) | |
| Gender (male) | 0.059 | 0.509 | 0.013 | 1 | 1.06 (0.39–2.87) | 0.908 |
| Age | −0.018 | 0.018 | 0.258 | 1 | 0.98 (0.94–1.01) | 0.324 |
| Living environment (urban) | 0.18 | 0.355 | 0.258 | 1 | 1.19 (0.59–2.4) | 0.611 |
| Education (college) | 1.923 | 0.749 | 6.6 | 1 | 6.84 (1.57–29.67) | |
| Marital status (with a partner) | −0.320 | 0.373 | 0.736 | 1 | 0.72 (0.34–1.5) | 0.391 |
| Family history of AUD | −0.023 | 0.377 | 0.004 | 1 | 0.97 (0.46–2.04) | 0.951 |
| Occupation (employed) | 0.093 | 0.365 | 0.065 | 1 | 1.09 (0.53–2.24) | 0.799 |
| Constant | −0.947 | 1.034 | 0.838 | 1 | 0.36 |
p-values were determined with logistic regression. Statistically significant p-values are bolded and marked with asterisk
To our knowledge, this study is the first to confirm the association between the rs1437396 polymorphism and AUD observed by Gelernter et al. (21). In addition, it provides preliminary evidence of even higher risk of developing depression comorbid with AUD associated with the T allele.
Our study has also singled out higher education as the only variable significantly associated with AUD and depression comorbidity (more frequent in university graduates). Literature data show the contrary – higher prevalence of depressive symptoms in people with low education (35). Furthermore, our findings about other environmental factors are not in line with evidenced association between depression and the employment, economic, and marital status, consumption of other psychoactive substances, cultural environment, and lifestyle (36–41). One possible explanation for this difference might be that in Romania, depressed people with lower education tend to avoid going to the doctor, especially to a psychiatrist due to the social stigma regarding mental illness. Most of the participants included in the study did not come to the hospital voluntarily. They were brought in by the ambulance as a medical emergency (violence in the context of alcohol intoxication or severe alcohol withdrawal symptoms).
Frequent AUD and MDD comorbidity can be explained by several theories. Alcohol has neurotoxic effects and negative social consequences (35), and some individuals use it as a form of self-medication for negative emotions (36). Genome-wide association studies (GWASs) have found multiple variants associated with AUD or MDD, but only one SNP has been associated with both disorders so far – the
The main limitation of this study is the small sample size, especially for the MDD subgroup. Furthermore, it does not explain how rs1437396 mediates AUD, and the best hypothesis is by modulating the effects of
In conclusion, this study has confirmed a significant association between AUD and the rs1437396 polymorphism and, for the first time, an even higher frequency of the T allele in the MDD subgroup of AUD patients. This result should be validated on a larger sample and perhaps shed some light on the genetic mechanisms behind this rather frequent comorbidity.