Increased alcohol consumption as a cause of alcoholism, without similar evidence for depression: A Mendelian randomization study

Marie Kim Wium-Andersen, David Dynnes Orsted, Janne Schurmann Tolstrup, Børge Grønne Nordestgaard

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

Resumé

BACKGROUND: Increased alcohol consumption has been associated with depression and alcoholism, but whether these associations are causal remains unclear. We tested whether alcohol consumption is causally associated with depression and alcoholism.

METHODS: We included 78 154 men and women aged 20-100 years randomly selected in 1991-2010 from the general population of Copenhagen, Denmark, and genotyped 68 486 participants for two genetic variants in two alcohol dehydrogenase (ADH) genes, ADH-1B (rs1229984) and ADH-1C (rs698). We performed observational and causal analyses using a Mendelian randomization design with antidepressant medication use and hospitalization/death, with depression and alcoholism as outcomes.

RESULTS: In prospective analyses, the multifactorially adjusted hazard ratio for participants reporting >6 drinks/day vs participants reporting 0.1-1 drinks/day was 1.28 (95% confidence interval, 1.00-1.65) for prescription antidepressant use, with a corresponding hazard ratio of 0.80 (0.45-1.45) for hospitalization/death with depression and of 11.7 (8.77-15.6) for hospitalization/death with alcoholism. For hospitalization/death with alcoholism, instrumental variable analysis yielded a causal odds ratio of 28.6 (95 % confidence interval 6.47-126) for an increase of 1 drink/day estimated from the combined genotype combination, whereas the corresponding multifactorially adjusted observational odds ratio was 1.28 (1.25-1.31). Corresponding odds ratios were 1.11 (0.67-1.83) causal and 1.04 (1.03-1.06) observational for prescription antidepressant use, and 4.52 (0.99-20.5) causal and 0.98 (0.94-1.03) observational for hospitalization/death with depression.

CONCLUSIONS: These data indicate that the association between increased alcohol consumption and alcoholism is causal, without similar strong evidence for depression.

OriginalsprogEngelsk
TidsskriftInternational Journal of Epidemiology
Vol/bind44
Udgave nummer2
Sider (fra-til)526-39
DOI
StatusUdgivet - 2015

Fingeraftryk

Random Allocation
Alcohol Drinking
Alcoholism
Depression
Alcohol Dehydrogenase
Odds Ratio
Prescriptions
Confidence Intervals
Denmark
Population

Citer dette

@article{8dcce287a2ca4a06abe17004b4bf3f8e,
title = "Increased alcohol consumption as a cause of alcoholism, without similar evidence for depression: A Mendelian randomization study",
abstract = "BACKGROUND: Increased alcohol consumption has been associated with depression and alcoholism, but whether these associations are causal remains unclear. We tested whether alcohol consumption is causally associated with depression and alcoholism.METHODS: We included 78 154 men and women aged 20-100 years randomly selected in 1991-2010 from the general population of Copenhagen, Denmark, and genotyped 68 486 participants for two genetic variants in two alcohol dehydrogenase (ADH) genes, ADH-1B (rs1229984) and ADH-1C (rs698). We performed observational and causal analyses using a Mendelian randomization design with antidepressant medication use and hospitalization/death, with depression and alcoholism as outcomes.RESULTS: In prospective analyses, the multifactorially adjusted hazard ratio for participants reporting >6 drinks/day vs participants reporting 0.1-1 drinks/day was 1.28 (95{\%} confidence interval, 1.00-1.65) for prescription antidepressant use, with a corresponding hazard ratio of 0.80 (0.45-1.45) for hospitalization/death with depression and of 11.7 (8.77-15.6) for hospitalization/death with alcoholism. For hospitalization/death with alcoholism, instrumental variable analysis yielded a causal odds ratio of 28.6 (95 {\%} confidence interval 6.47-126) for an increase of 1 drink/day estimated from the combined genotype combination, whereas the corresponding multifactorially adjusted observational odds ratio was 1.28 (1.25-1.31). Corresponding odds ratios were 1.11 (0.67-1.83) causal and 1.04 (1.03-1.06) observational for prescription antidepressant use, and 4.52 (0.99-20.5) causal and 0.98 (0.94-1.03) observational for hospitalization/death with depression.CONCLUSIONS: These data indicate that the association between increased alcohol consumption and alcoholism is causal, without similar strong evidence for depression.",
author = "Wium-Andersen, {Marie Kim} and Orsted, {David Dynnes} and Tolstrup, {Janne Schurmann} and Nordestgaard, {B{\o}rge Gr{\o}nne}",
note = "{\circledC} The Author 2014; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association.",
year = "2015",
doi = "10.1093/ije/dyu220",
language = "English",
volume = "44",
pages = "526--39",
journal = "International Journal of Epidemiology",
issn = "0300-5771",
publisher = "Heinemann",
number = "2",

}

Increased alcohol consumption as a cause of alcoholism, without similar evidence for depression : A Mendelian randomization study. / Wium-Andersen, Marie Kim; Orsted, David Dynnes; Tolstrup, Janne Schurmann; Nordestgaard, Børge Grønne.

I: International Journal of Epidemiology, Bind 44, Nr. 2, 2015, s. 526-39.

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

TY - JOUR

T1 - Increased alcohol consumption as a cause of alcoholism, without similar evidence for depression

T2 - A Mendelian randomization study

AU - Wium-Andersen, Marie Kim

AU - Orsted, David Dynnes

AU - Tolstrup, Janne Schurmann

AU - Nordestgaard, Børge Grønne

N1 - © The Author 2014; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association.

PY - 2015

Y1 - 2015

N2 - BACKGROUND: Increased alcohol consumption has been associated with depression and alcoholism, but whether these associations are causal remains unclear. We tested whether alcohol consumption is causally associated with depression and alcoholism.METHODS: We included 78 154 men and women aged 20-100 years randomly selected in 1991-2010 from the general population of Copenhagen, Denmark, and genotyped 68 486 participants for two genetic variants in two alcohol dehydrogenase (ADH) genes, ADH-1B (rs1229984) and ADH-1C (rs698). We performed observational and causal analyses using a Mendelian randomization design with antidepressant medication use and hospitalization/death, with depression and alcoholism as outcomes.RESULTS: In prospective analyses, the multifactorially adjusted hazard ratio for participants reporting >6 drinks/day vs participants reporting 0.1-1 drinks/day was 1.28 (95% confidence interval, 1.00-1.65) for prescription antidepressant use, with a corresponding hazard ratio of 0.80 (0.45-1.45) for hospitalization/death with depression and of 11.7 (8.77-15.6) for hospitalization/death with alcoholism. For hospitalization/death with alcoholism, instrumental variable analysis yielded a causal odds ratio of 28.6 (95 % confidence interval 6.47-126) for an increase of 1 drink/day estimated from the combined genotype combination, whereas the corresponding multifactorially adjusted observational odds ratio was 1.28 (1.25-1.31). Corresponding odds ratios were 1.11 (0.67-1.83) causal and 1.04 (1.03-1.06) observational for prescription antidepressant use, and 4.52 (0.99-20.5) causal and 0.98 (0.94-1.03) observational for hospitalization/death with depression.CONCLUSIONS: These data indicate that the association between increased alcohol consumption and alcoholism is causal, without similar strong evidence for depression.

AB - BACKGROUND: Increased alcohol consumption has been associated with depression and alcoholism, but whether these associations are causal remains unclear. We tested whether alcohol consumption is causally associated with depression and alcoholism.METHODS: We included 78 154 men and women aged 20-100 years randomly selected in 1991-2010 from the general population of Copenhagen, Denmark, and genotyped 68 486 participants for two genetic variants in two alcohol dehydrogenase (ADH) genes, ADH-1B (rs1229984) and ADH-1C (rs698). We performed observational and causal analyses using a Mendelian randomization design with antidepressant medication use and hospitalization/death, with depression and alcoholism as outcomes.RESULTS: In prospective analyses, the multifactorially adjusted hazard ratio for participants reporting >6 drinks/day vs participants reporting 0.1-1 drinks/day was 1.28 (95% confidence interval, 1.00-1.65) for prescription antidepressant use, with a corresponding hazard ratio of 0.80 (0.45-1.45) for hospitalization/death with depression and of 11.7 (8.77-15.6) for hospitalization/death with alcoholism. For hospitalization/death with alcoholism, instrumental variable analysis yielded a causal odds ratio of 28.6 (95 % confidence interval 6.47-126) for an increase of 1 drink/day estimated from the combined genotype combination, whereas the corresponding multifactorially adjusted observational odds ratio was 1.28 (1.25-1.31). Corresponding odds ratios were 1.11 (0.67-1.83) causal and 1.04 (1.03-1.06) observational for prescription antidepressant use, and 4.52 (0.99-20.5) causal and 0.98 (0.94-1.03) observational for hospitalization/death with depression.CONCLUSIONS: These data indicate that the association between increased alcohol consumption and alcoholism is causal, without similar strong evidence for depression.

U2 - 10.1093/ije/dyu220

DO - 10.1093/ije/dyu220

M3 - Journal article

C2 - 25433705

VL - 44

SP - 526

EP - 539

JO - International Journal of Epidemiology

JF - International Journal of Epidemiology

SN - 0300-5771

IS - 2

ER -