www.jech.com
574 Hmlinen, Kaprio, Isomets, et al
www.jech.com
Cigarette smoking, alcohol intoxication and major depressive episode 575
1.82), and three or more chronic diseases (OR is also possible that smoking may be a marker
2.49, 95%CI 1.85, 3.34). We found no of a variety of genetic, personal, social and
evidence for interaction eVects between alcohol familial properties.26 Thus, resolution of cau-
and smoking in the logistic regression analyses sality even in longitudinal studies is not a sim-
(table 2). ple issue.
On the population level daily tobacco use However nicotine, the pharmacologically
(over 10 cigarettes/day) in this model was a active component of cigarette smoke, is known
very important contributor to major depressive to have direct and indirect eVects on the
episode (APT = 0.11), although less so than neurotransmitters thought to be involved in
having three or more chronic conditions (APT major depression.27 There is speculation of a
= 0.16). Alcohol intoxication had a smaller specific relation between smoking and major
impact on the population level (APT = 0.03) depression only in smokers who are nicotine
because its prevalence is low despite a higher dependent.26 This hypothesis is suggested by
relative risk. findings2 indicating that increased rates of
major depression are found only among
Discussion nicotine dependent smokers and not non-
We found that alcohol intoxication at least once dependent smokers. Our findings suggest a
per week and smoking 10 or more cigarettes relation between the prevalence of major
daily is associated with the risk of major depressive episode and the consumption of
depressive episode. This study also reports that cigarettes: current smoking over 10 cigarettes
at the population level both cigarette smoking per day was significantly associated with major
and alcohol intoxication, after adjusting for depressive episode. The increased association
other risk factors, are associated with major for those smoking under 10 cigarettes per day
depressive episode independently of each or for irregular smokers was less consistent.
other. Among former smokers, the prevalence of
This study involved a large random sample major depression was increased only in
of a non-institutionalised general population women.
(n=5993) representing the whole of Finland. The increased incidence of alcoholism
The participation rate of households inter- among smokers16 may be partially responsible
viewed was high and the interviews were made for the relation observed between smoking and
by professional interviewers specially trained depression, although some such findings have
by two experienced psychiatrists in the use of been independent of alcohol intake.13 Factors
the UM-CIDI Short Form. The study sample such as education may also exert a rolelower
is closely comparable to the entire Finnish educational attainment is often associated with
population for the sociodemographic charac- a greater prevalence of smoking28 29 and has
teristics investigated.19 The Short Form of the been implicated in failure to stop.28 30 Other
UM-CIDI is a shortened form plus an suggested confounding factors between smok-
algorithm based on the NCS findings. The ing and depression include age,31 obesity,32 and
characteristics of this method are discussed in sex, with many studies showing a stronger rela-
the FHCS prevalence study.19 No indication of tion between smoking and depressive symp-
cultural biases in the validity of UM-CIDI has toms among women.4 3335 In our model the
arisen. Our prevalence estimate for major other possible confounding factors were ad-
depressive episode was convergent with most justed for one by one, after which only frequent
other general population studies.19 Our instru- alcohol intoxication (weekly or more often) and
ment did not generate diagnoses of specific sex remained statistically significant in the final
mood disorders but only of major depressive model. Neither educational attainment, obesity
episode. It is probable that most of the people nor age had a statistically significant eVect.
with a major depressive episode had unipolar Most studies investigating the comorbidity
major depression. For comparison, in the Epi- between major depression and alcoholism have
demiological Catchment Area (ECA) study23 focused on diagnostic level information36 37; few
77% of subjects with major depressive episode have assessed the nature of relations between
had unipolar major depression and most of the depressive symptoms and problematic alcohol
remainder were suVering from bipolar disor- use.38
ders or bereavement. We also lacked diagnostic In conclusion, both cigarette smoking and
level information on substance abuse or alcohol intoxication seem to be important risk
dependence. In earlier studies with adults, the factors for major depressive episode. In our
association between smoking and MDD,24 and population the impact of smoking was greater.
(mild or moderate) nicotine dependence and
MDD,25 remained significant even after con- Funding: the study was financially supported by the Yrj Jahns-
son Foundation and the Academy of Finland (grant number
trolling for other psychiatric disorders. 42044).
These cross sectional data can only suggest, Conflicts of interest: none.
not prove, causal relations. The strong associ-
ation of cigarette smoking with major depres- 1 Glassman AH, Helzer JE, Covey LS, et al. Smoking, smok-
ing cessation, and major depression. JAMA
sion indicates that smoking may be an 1990;264:15469.
important risk factor both on the individual 2 Breslau N, Kilbey MM, Andreski P. Nicotine dependence,
major depression, and anxiety in young adults. Arch Gen
and population level. It has been suggested that Psychiatry 1991;48:106974.
causes of major depression may include both 3 Kendler KS, Neale MC, MacLean CJ, et al. Smoking and
major depression: a causal analysis. Arch Gen Psychiatry
genetic and enviromental factors.24 There may 1993;50:3643.
be separate specific causal mechanisms from 4 Isomets E, Aro S, Aro H. Depression in Finland: a compu-
ter assisted telephone interview study. Acta Psychiatr Scand
smoking to major depression, and vice versa. It 1997;96:1228.
www.jech.com
576 Hmlinen, Kaprio, Isomets, et al
5 Hughes JR, Hatsukami DK, Mitchell JE, et al. Prevalence of 24 Kendler KS, Neale MC, MacLean CJ, et al. Smoking and
smoking among psychiatric outpatients. Am J Psychiatry major depression. A causal analysis. Arch Gen Psychiatry
1986;14:9937. 1993;50:3643.
6 Glassman AH, Stetner F, Walsh T, et al. Heavy smokers, 25 Breslau N, Kilbey M, Andreski P. Nicotine dependence,
smoking cessation, and clonidine. JAMA 1988;259:28636. major depression, and anxiety in young adults. Arch Gen
7 Kandel DB, Davies M. Adult sequelae of adolescent depres- Psychiatry 1991;48:106974.
sive symptoms. Arch Gen Psychiatry 1986;43:25562.
8 Anda RF, Williamson DF, Escobedo LG, et al. Depression 26 Breslau N, Peterson EL, Schultz LR, et al. Major depression
and the dynamics of smoking. JAMA 1990;264:15415. and stages of smoking. Arch Gen Psychiatry 1998;55:1616.
9 Patton GC, Hibbert M, Rosier MJ, et al. Is smoking associ- 27 Hall SM, Munoz RF, Reus VI, et al. Nicotine, negative
ated with depression and anxiety in teenagers? Am J Public aVect, and depression. J Consult Clin Psychol 1993;61:761
Health 1996; 86:22530. 7.
10 Covey LS, Hughes DC, Glassman AH, et al. Ever-smoking, 28 Pierce JP, Fiore MC, Novotny TE, et al. Trends in cigarette
quitting, and psychiatric disorders: evidence from the Dur- smoking in the United States: educational diVerences are
ham, North Carolina, Epidemiologic Catchment Area. increasing. JAMA 1989;261:5660.
Tobacco Control 1994;3:2227. 29 Wagenknecht LE, Perkins LL, Cutter GR, et al. Cigarette
11 Glassman AH, Stetner F, Walsh BT, et al. Heavy smokers, smoking behaviour is strongly related to educational status:
smoking cessation, and clonidine. Results of a double-
blind, randomized trial. JAMA 1988;259:28636. the CARDIA study. Prev Med 1990;19:15869.
12 Merikangas K, Gelernter C. Comorbidity for alcoholism 30 Hatziandreu EJ, Pierce JP, Lefkopoulou M, et al. Quitting
and depresssion. Psychiatr Clin North Am 1990;13:61332. smoking in the United States in 1986. J Natl Cancer Inst
13 Helzer JE, Przybeck T. The co-occurence of alcoholism with 1990;82:14026.
other psychiatric disorders in the general population and its 31 Maier W, Lichtermann D, Oehrlein A, et al. Depression in
impact on treatment. J Stud Alcohol 1988;49:21924. the community: a comparison of treated and non-treated
14 Aneshensel C, Huba G. Depression, alcohol use, and smok- cases in two non-referred samples. Psychopharmacology
ing over one year: a four-wave longitudinal causal model. J 1992;106 (suppl):7981.
Abnorm Psychol 1983;92:13450. 32 Istvan J, Zavela K, Weidner G. Body weight and psychologi-
15 Covey LS, Glassman AH, Stetner F, et al. EVect of history of cal distress in NHANES I. Int J Obes Relat Metab Disord
alcoholism or major depression on smoking cessation. Am 1992;16:9991003.
J Psychiatry 1993;150:15467. 33 Frederick T, Frerichs RR, Clark VA. Personal health habits
16 DiFranza JK, Guerrera MP. Alcoholism and smoking. J and symptoms of depression at the community level. Prev
Stud Alcohol 1990;51:1305.
17 Glassman AH. Blue mood, blackened lungs:depression and Med 1988;17:17382.
smoking. JAMA 1990;264:15834. 34 Peres-Stable EJ, Marin G, Marin BV, et al. Depressive
18 Glassman AH, Covey LS, Dalack GW, et al. Smoking cessa- symptoms and cigarette smoking among Latinos in San
tion, clonidine, and vulnerability to nicotine among Francisco. Am J Public Health 1990;80:15002.
dependent smokers. Clin Pharmacol Ther 1993;54:6709. 35 Lee DJ, Markides KS. Health behaviours, risk factors, and
19 Lindeman S, Hmlinen J, Isomets E, et al. The 12-month health indicators associated with cigarette use in Mexican
prevalence and risk factors for major depressive episode in Americans: results from the Hispanic HANES. Am J Public
Finland: representative sample of 5993 adults. Acta Psychi- Health 1991;81:85964.
atr Scand 2000;102:17884. 36 Bromet EJ, Parkinson DK, Curtis EC, et al. Epidemiology of
20 Arinen S, Hkkinen U, Klaukka T, et al. Health and use of depression and alcohol abuse/dependence in a managerial
health services in Finland. Main findings of the Finnish Health and professional work force. J Occup Med 1990;32:98995.
Care Survey and changes from 1987. [In Finnish, English 37 Kessler RC, Mc Gonagle KA, Zhao S, et al. Lifetime and
summary). Helsinki: OYcial Statistics of Finland, 1998.
21 Liang KY, Zeger SL. Longitudinal data analysis using gen- 12-month prevalence of DSM-III-R psychiatric disorders
eralized linear models. Biometrika 1986;73:1322. in the United States: a result from the National Comorbid-
22 Rothman KJ. Modern epidemiology. Boston: Little, Brown, ity Survey. Arch Gen Psychiatry 1994;51:818.
1989. 38 Moscato BS, Russell M, Zielezny M, et al. Gender
23 Weissman MM, Bruce M, Leaf P, et al. AVective disorders. diVerences in the relation between depressive symptoms
In: Robins L, Regier D, eds. Psychiatric disorders in America. and alcohol problems: a longitudinal perspective. Am J
New York: Free Press 1991:5380. Epimediol 1997;11:96674.
www.jech.com