The authors have declared that no competing interests exist.
Conceived and designed the experiments: RP PK MW IC SF MF. Performed the experiments: IC KG KS. Analyzed the data: IC PS RP. Contributed reagents/materials/analysis tools: GB PK AP SF MF. First draft of the manuscript: IC RP. Critical reading and comments: IC KG KS GB SF MF PS AP PK MW RP.
rs6943555 in
We genotyped rs6943555 in 625 completed suicides and 3861 controls using real-time TaqMan Allelic Discrimination Assay. All individuals were Polish Caucasians.
We detected an association between suicide and rs6943555 A allele (
In our study rs6943555 A allele is associated with suicide committed after drinking ethanol shortly before death. The rs6943555 A allele may be linked to adverse emotional reaction to ethanol, which could explain the association with lower consumption in general population as well as the predisposition to suicide under influence of ethanol.
Suicidal behavior encompasses a spectrum ranging from the suicidal thoughts to suicide attempts and completed suicide
Recently rs6943555 in
There are numerous reports linking alcohol abuse and the risk for suicide
Given these data we hypothesized that there was an association between rs6943555 and completed suicide, which could possibly be modulated by exposure to ethanol shortly before death. The aim of out study was to test this hypothesis.
Suicide victims were consecutive cases from the Warsaw metropolitan area autopsied in the Department of Forensic Medicine at the Medical University of Warsaw, Poland. The information about clinical variables, such as age, gender, suicide method, blood ethanol concentration, psychiatric diagnosis and history of addiction was collected from the post-mortem medical and forensic examination protocols. Due to limited number of cases, for which definite negative history of addiction and information about psychiatric diagnosis could be obtained, we pooled those with the negative history with the cases, for which data were missing. Characteristics of suicides are shown in
Number of cases (%) | ||
Males | 494 (81.3) | |
Females | 114 (18.7) | |
mean = 43.71, SD = 16.82 | 542 | |
Under influence of alcohol (ethanol concentration > 20 mg/dl), mean = 185.0, SD = 110.1 | 235 (37.6) | |
Without influence of alcohol (ethanol concentration ≤ 20 mg/dl) | 390 (62.4) | |
Hanging | 446 (76.6) | |
Jumping from a high place | 68 (11.7) | |
Self-harm by sharp object | 17 (2.9) | |
Shot with a firearm | 17 (2.9) | |
Jumping or lying before moving object | 14 (2.4) | |
Toxic effect from ingested substance | 11 (1.9) | |
Other | 9 (1.6) | |
|
No | 585 (93.6) |
Yes: | 40 (6.4) | |
Depression | 13 (32.5) | |
Schizophrenia | 2 (5.0) | |
Other or unknown | 25 (62.5) | |
|
No | 602 (96.3) |
Yes: | 23 (3.7) | |
Alcoholism | 21 (91.3) | |
Drug addiction | 1 (4.4) | |
Alcoholism and drug addiction | 1 (4.4) |
We observed statistically significant difference regarding gender between cases and controls (47.11% males for controls,
We observed statistically significant difference regarding age between cases and controls (mean = 45.70, SD = 14.91 for controls,
Number of cases (%) | ||
Males | 117 (89.3) | |
Females | 14 (10.7) | |
mean = 42.89, SD = 16.76 | 122 | |
Under influence of alcohol (ethanol concentration > 20 mg/dl), mean = 157.5, SD = 86.5 | 65 (52.4) | |
Without influence of alcohol (ethanol concentration ≤ 20 mg/dl) | 59 (47.6) | |
Hanging | 101 (79.5) | |
Jumping from a high place | 14 (11.0) | |
Self-harm by sharp object | 3 (2.4) | |
Shot with a firearm | 2 (1.6) | |
Jumping or lying before moving object | 5 (3.9) | |
Toxic effect from ingested substance | 1 (0.8) | |
Other | 1 (0.8) | |
|
No | 111 (84.1) |
Yes: | 21 (15.9) | |
Depression | 11 (52.4) | |
Schizophrenia | 2 (9.5) | |
Other or unknown | 8 (38.1) | |
|
No | 127 (96.2) |
Yes: | 5 (3.8) | |
Alcoholism | 5 (100) | |
Drug addiction | 0 | |
Alcoholism and drug addiction | 0 |
A special consideration was given to the use of postmortem samples from suicides from whom informed consent could not be obtained. The testing was performed using a part of blood sample routinely collected during all autopsies. All samples and phenotype data were anonimised immediately after obtaining. An approval of this protocol was obtained from the Warsaw Medical University Ethical Board (KB/185/2010).
The control group, which has been used in a previous study
The alcohol drinking habits, including the size and frequency of drinks, were obtained from a questionnaire administered by a certified interviewer as described previously
Vertical bars indicate 95% confidence intervals.
All cases and controls were Polish Caucasians. The study was approved by the Bioethical Committee of Medical University of Warsaw and all control subjects gave written consent for the anonymous use of their DNA for research.
Genomic DNA from cases and from controls was isolated from EDTA blood samples using salting out procedure described by Miller
We successfully genotyped 625 completed suicide subjects (+ 62 undetermined samples, 91% call rate) and 3861 controls (+ 157 undetermined genotypes, 96% call rate). Our study could detect with power of 0.8 (alpha = 0.05) an allelic association conferring OR = 1.17.
Distribution of rs6943555 genotypes among cases and controls was compared assuming dominant, co-dominant or recessive models of effect using Web-Assotest program (
Comparison of the genotype distribution after adjusting for age and gender was performed by multiple logistic regression analysis using SPSS software Release 11.5.0.
Search for genotype-phenotype correlation among suicides was performed by chi square test or Mann-Whitney U test using Statistica software version 10.0 (StatSoft, Inc., Tulsa, OK).
The distribution of genotypes was in Hardy-Weinberg Equilibrium (HWE) among subjects and controls (
Cohort | rs6943555 genotypes | HWE | Model | |||||
Allelic comparison | Recessive | Co−dominant | Dominant | |||||
OR (CI), |
OR (CI) |
OR (CI) |
OR (CI) |
|||||
TT (%) | AT (%) | AA (%) | (AA vs. TT/AT) | (AA vs. AT vs. TT) | (AT/AA vs. TT) | |||
Suicide victims I ( |
359 (57.4) | 232 (37.1) | 34 (5.4) | 0.661 | 1.171 (1.03; 1.37), 0.018 | 1.161 (0.79; 1.68), 0.456, 0.018 | 1.181 (1.03; 1.36), 0.020, 0.381 | 1.241 (1.05; 1.48), 0.013, 0.970 |
1.192 (1.03; 1.38), 0.017 | 1.172 (0.80; 1.73), 0.426, 0.019 | 1.192 (1.03; 1.38), 0.019, 0.416 | 1.252 (1.05; 1.49), 0.013, 0.924 | |||||
Suicide victims I with blood ethanol concentration > 20 mg/dl ( |
126 (53.6) | 92 (39.1) | 17 (7.2) | 0.971 | 1.381 (1.11; 1.70), 0.003 | 1.571 (0.94; 2.62), 0.105, 0.017 | 1.371 (1.11; 1.69), 0.004, 0.770 | 1.451 (1.12; 1.89), 0.006, 0.396 |
Suicide victims I with blood ethanol concentration ≤ 20 mg/dl ( |
233 (59.7) | 140 (35.9) | 17 (4.36) | 0.482 | 1.081 (0.90; 1.29) 0.405 | 0.921 (0.55; 1.52), 0.732, 0.200 | 1.081 (0.90; 1.28), 0.411, 0.298 | 1.131 (0.92; 1.40), 0.256, 0.493 |
Suicide victims II ( |
73 (54.5) | 51 (38.1) | 8 (6.0) | 0.818 | 1.28 (0.96–1.69) 0.089 | 1.30 (0.62–2.69) 0.50, 0.12 | 1.27 (0.96–1.68) 0.099, 0.064 | 1.36 (0.96–1.92) 0.089, 0.85 |
Suicide victims II with blood ethanol concentration > 20 mg/dl ( |
31 (47.7) | 30 (46.2) | 4 (6.2) | 0.352 | 1.551 (1.06–2.27) 0.023 | 1.321 (0.47–3.66) 0.61, 0.017 | 1.541 (1.05–2.25) 0.031, 0.26 | 1.841 (1.13–3.01) 0.015, 0.9 |
Suicide victims II with blood ethanol concentration ≤ 20 mg/dl ( |
39 (66.1) | 16 (27.1) | 4 (6.8) | 0.210 | 0.961 (0.61–1.51) 0.85 | 1.461 (0.52–4.08) 0.50, 0.42 | 1 (0.61–1.50) 0.86, 0.30 | 0.861 (0.50–1.48) 0.60, 0.36 |
Controls ( |
2420 (62.7) | 1258 (32.6) | 183 (4.7) | 0.236 | ||||
Controls BDI ≤ 11 ( |
1681 (62.8) | 870 (32.5) | 125 (4.7) | 0.362 |
P values <0.05 were
The association between suicide and rs6943555 remained significant also after adjusting for age and gender (OR = 1.29,
Given the previously reported relationship between rs6943555 and alcohol consumption
We did not find statistically significant differences in the distribution of rs6943555 genotypes among suicide after stratifying for gender (
Relatively many samples from suicides (62 or 9%) had low DNA quality which precluded genotyping. This was most likely caused by low initial quality of some samples and their long storage (>5 years in some cases). Since high genotype failure rate may introduce a bias due to preferential genotyping of one allele we attempted a replication of our results using a second group of samples (
The distribution of rs6943555 genotypes in the whole second group of cases as well as after stratifying according to blood ethanol concentration is shown in
When we pooled all suicides with blood ethanol concentration >20 mg/dl (
The main novel finding from our study is the association between the A allele of rs6943555 in the
The direction of the observed association of the rs6943555 A allele with suicide is opposite to the association with alcohol consumption reported previously
Although we show that rs6943555 is associated with suicide it is not clear whether this is a primary association or an effect of marker(s) in linkage disequilibrium. It is also possible that multiple variants in
It is increasingly recognized that pathophysiology of mental disorders is heterogenic and multifactorial
The authors would like to thank Waldemar Szelenberger, MD, PhD for his support for this study.