The Alcohol-Depression Hypothesis

The alcohol-depression hypothesis stems from the observation that a raised  prevalence of depression occurs in groups in which alcohol use is low. The hypothesis has to be reconciled with the well-established covariance of depressive disorders with alcohol abuse.  The alcohol-depression hypothesis suggests that up to a point, alcohol use masks or inhibits depression. It also implies that cultures which support alcohol use will also support beliefs about the benefits of alcohol use, including mental health benefits. Conversely, alcohol use will be seen as dangerous and bad in cultures which do not support alcohol use. Evidence in support of the hypothesis is reviewed, chiefly from research focusing on Jewish groups. This research generally shows strong culturally-carried beliefs which do not tolerate recreational drinking or drunkenness, and which do not cover any possible mental health benefits of alcohol use. These beliefs go alongside low levels of alcohol consumption. Gender differences in drinking patterns are also considered. Other interpretations of the material are considered, and inconsistencies and gaps in research are identified

INTRODUCTION

In the early 1990s, we were analysing and writing up a study of life-events and depression in the orthodox Anglo-Jewish community (1). We had been carefully trained in the use of life-events methodology and psychiatric diagnosis, and the analyses were generally showing the expected relationships between stress –  assessed from life-events and difficulties interviews (LEDS) (2) - and minor psychiatric disorder including depression. But there was some alarm in the research team. Women were supposed to be twice as likely to suffer from depression as were men, but our data were clearly showing an almost identical prevalence of unipolar depression among men as among women. What was wrong with our data? Was it our methodology, our sample, our analyses? We scrutinised everything carefully, and finally came to the conclusion that perhaps our methods, sample and analyses were good enough. We wondered if  there was some connection between the low levels of alcohol use in the Jewish community, and the raised prevalence of depression among the men. We became aware of other reports citing similar findings among Jews (3, 4). 

We tentatively suggested that low alcohol use might be a factor explaining the raised prevalence of depression among men, and went on to compare the attitudes and beliefs associated with alcohol use among Jews and Protestants in the UK (5, 6). This chapter explores the current status of the hypothesis that low use of alcohol may raise the prevalence of depression – the alcohol-depression hypothesis.

THE RELATIONS BETWEEN DEPRESSION AND ALCOHOL USE

The alcohol-depression  hypothesis suggests an inverse relationship between the prevalence of depression and alcohol use. The hypothesis also implies that beliefs about the effectiveness of alcohol as a mood enhancer will co-vary with alcohol use,being less favourable when alcohol use is very limited..

We know that the reported relations between depression and alcohol use are paradoxical. On the one hand it is well established that the prevalence of major depressive disorder and alcohol abuse disorders covary (7, 8). One possible explanatory factor may be applicable in explaining this effect, as well as the apparent paradox  that groups which do not tolerate  use of alcohol have a raised prevalence of major depression.

Alcohol may be used as self-medication for depression and other stress-induced negative states (notably anxiety and anger). (7, 8, 9) This may lead to an out-of-control spiralling of use, and eventual abuse. There is some evidence that depression onset may precede the abuse of alcohol and (other) substances (8). Self-medication may be a mixed blessing. Moderate alcohol consumption may induce feelings of well-being, but even moderate alcohol consumption may be followed by depressed mood and associated changes in serotonin (10). Heavy alcohol consumption has many unpleasant physical and psychological consequences, as well as depressed mood sequelae. Thus alcohol abuse can be both an effect and a cause of depressive disorder.

The sequence of events would be:

Stress - depression – alcohol use – raised mood followed by further depression – further alcohol use – alcohol abuse - financial problems, unacceptable social behaviour – further depression

This sequence of events is consistent with the finding that alcohol abuse covaries with depression. 

It leaves us with the apparently paradoxical finding that in groups with high prevalence of depression, there are low rates of alcohol dependency and abuse. Thus an analysis of national US data (11) found that rates of alcohol abuse/dependence were inversely related to rates of major depression. Jewish males had higher rates of major depression and lower rates of alcohol abuse (at a ratio of 4.5:1), whereas non-Jewish males had lower rates of major depression and higher rates of alcohol abuse (a ratio of 0.4:1). Also, rates of depression in male Jews were found to be reduced when they were living in areas of greater assimilation and had adopted local behaviours such as increased alcohol consumption. While in most groups, depression prevalence is approximately double among women compared to men, a UK study reported that major depressive disorder was as prevalent among orthodox Jewish men as among women (1). This study suggested that a contributory factor may have been the zero reported prevalence of alcohol abuse. Another UK study suggested that high rates of depression within Jewish men living in London may be explained in terms of their lower use of alcohol, compared to other men (3). Gender equality in rates of depression has also been found among Jews in Israel (4), within the Amish community in the USA (12), and among diabetics(13). However evidence on depression prevalence in Muslim countries – in which alcohol consumption is normally illegal - has not given much support to the expectation that low alcohol use will be associated with raised depression, particularly in men. One study has reported similar depression prevalence among men and women, among  the Turkish elderly (14), but in this study depression was related to alcohol consumptiom. Other studies in countries which prohibit alcohol consumption have indicated higher depression among women than among men. (15, 16). An important systematic review of 17 studies estimating depression prevalence in Pakistan (17) showed  very high prevalences of depression among both men and women, with overall mean prevalence of mixed depression and anxiety among women approximately double (29-66%) that among men (10-33%). Individual studies included in this review reported highr prevalence of depression among women compared to men. These reports indicate the major extent to which factors others than alcohol use are related to depression. Thus the two findings in Arab countries (15, 16) indicated the impact of financial hardship, ill-health and other adversity in depression, all of which may fall more heavily on women whose economic and social positions may be very weak compared to that of women in Western countries.  The systematic review of studies in Pakistan included seven analyses of life-events and difficulties: economic difficulties and poor education, being a housewife, relationship problems with in-laws, and absence of confiding supportive relationships were factors associated with depression. This confirms the suggestion that women’s weak economic and social position in Pakistan is a key factor in depression. Throughout the South Indian subcontinent it is normative for a woman to become the property of her husband’s family, having little or no contact with her family of origin. These data attest to the extent to which women are highly dependent on the goodwill of their husbands’ families. These studies in Muslim countries also attest to the extent to which the effects of negligible alcohol use on depression prevalence are heavily outweighed by other factors.

If alcohol use does affect depression, and if we assume that alcohol abuse is the tip of an iceberg in a society in which alcohol use is widespread, we could suggest that although alcohol abuse has spiralling deleterious effects on mental health, if alcohol consumption remains moderate, the benefits may outweigh the mental health risks. The first benefit is mood improvement: feelings of well-being and euphoria, and lessening of negative moods have all been reported (6). Secondly, alcohol consumption in social settings facilitates conviviality, conversation and confiding. These include aspects of social support whose mental health benefits have been well-documented (2, 18). Thus it is suggested that moderate alcohol use/recreational drinking may be effective means of coping, and are certainly seen as such by lay individuals (6). Moderate alcohol use, particularly in the context of recreational and social drinking, may  reduce the risk of depression, or mask the symptoms of depression. It is unlikely to carry along other risks of heavy drinking: financial overburdening, and unacceptable social behaviour. Moderate self-medication with alcohol may be so moderate that the depressing morning-after effects and other unpleasant effects are minimal, and  mood-enhancing and social support benefits are maximal.

The sequence of events would be: 

Stress – depression – alcohol use – raised mood plus social support

This hypothesis is contingent on the existence of a body of moderate alcohol users, successfully self-medicating. The most compelling evidence is offered by a study of the Los Angeles Catchment Area cohort (9). In the presence of stress (chronic strain and negative life events) there was U-shaped relationship between depression and alcohol consumption.  Those with the highest depression scores when stressed were heavy drinkers, light drinkers and non-drinkers. The lowest depression scores were reported by moderate and light-moderate drinkers. Lipton suggests that moderate alcohol use may be one of several behaviours that either suppress or attenuate the effects of stress on depression. Other work has confirmed the U-shaped relationship between alcohol consumption and health, for example it has been reported that moderate alcohol consumption reduces the risk of coronary heart disease, diabetes and dementia (19, 20, 21)

ALCOHOL USE AND ATTITUDES  

As a corollary of the alcohol-depression hypothesis, we would expect alcohol consumption to be zero or low, and social attitudes and beliefs about alcohol consumption to be negative in those groups with a raised prevalence of depression among men. 

Although the evidence is patchy, there is some consistency with this expectation. 

What do we know about prescribed amounts of alcohol consumption, and how do these relate to “light”, “moderate” and “heavy” drinking?

One or two drinks a day (perhaps up to four for men), five to six days a week has been specified as moderate consumption (19). Although there is evidence that moderate consumption may have protective health benefits, there is widespread reluctance to recommend moderate alcohol consumption as a prophylactic measure. This is because the protective mechanisms are poorly understood and the risks of a spiral into excessive alcohol use are too great.

In Judaism, a small vessel of wine (about 100ml) is normally required 3 times a week for religious ceremonies. On some of these occasions, grape juice, spirits or tea may used as alternatives. On certain festivals, slightly larger amounts are required – for example four cups of wine of Passover, and actual drunkenness on Purim. These ceremonial occasions for alcohol consumption are more obligatory for men than they are for women and children. There is no history of teetotalism as a desirable religious goal within Judaism. 

Staunch minorities in the Christian tradition have opposed the use of alcohol. For example the teetotal movement, embedded in Christianity, arose to stem the horrific social and health problems caused by alcohol in the rapidly urbanising cities of the Western world during the C19th. Nevertheless Christianity and Christian societies have been broadly tolerant of alcohol use, and meeting-places for the consumption of alcohol (bars, inns, public houses) are important social centres in Christian countries. Alcohol consumption is on average higher among Christian groups compared to Jewish groups.

In Islam, alcohol sale and use are prohibited in many Muslim countries, and normatively, Muslims disapprove of its use (26, 27), though wine drinking is mentioned in the Koran. It has been reported that while very few Muslims use alcohol, when alcohol is easily available, alcohol use is more likely to spiral into abuse among those few (usually men) who do (27). 

For diabetics, zero or minimal alcohol consumption is normally recommended, though there is some evidence that consumption at moderate levels may be harmless or beneficial (21).

Given the available information about what is considered low, moderate and high consumption of alcohol, the information on religiously-prescribed drinking does suggest that among observant Jews and Muslims (and in Christian groups recommending abstinence), alcohol consumption is likely to be negligible or light, and there is some evidence supporting this (26, 27). Figure 1 shows the average frequency of alcohol consumption among 70 UK Jews and 91 people of Protestant affiliation and background (5). The range is from 0 (not at all) to 5 (every day). Figure 2 shows the average number of drinks consumed on each occasion, ranging from 0 (nothing) to 4 (4 or more). 

Figure 1

Figure 2

These data suggest that the men of Protestant affiliation and background were generally consuming moderate amounts of alcohol, as defined above, while Protestant women, and all Jews, were reporting light alcohol consumption.

As for attitudes to alcohol use, historically, Christianity and Judaism have been ambivalent to alcohol and its effects. This ambivalence has a biblical basis. Alcohol can improve mood (e.g.” Wine gladdens the heart of man”, Psalm 104, v15) , but can also lead to a loss of control and a range of undesirable behaviours (e.g. “And (Noah) drank from the wine and was drunk and was uncovered” Genesis 9, 21)

. 

The most detailed evidence about contemporary attitudes to alcohol use comes from investigations of Jewish groups, sometimes involving comparisons with Christian groups. Jews are less likely than other groups in western countries to use and abuse alcohol (22, 23, 24, 25). It has been suggested that one reason for this is their vulnerable and marginal position in society, making it important to remain sober and stay in control. In Israel, Jews regarded alcohol consumption as more dangerous than did Christians (26). They were less likely to agree that alcoholics could taper off and control their drinking again, and they were more likely to agree that alcoholics had to quit forever in order to recover. A UK study reported lighter alcohol consumption among Jews than among those of Protestant background, and this went along with beliefs and attitudes that were unaccepting of alcohol use. Jews saw drunkenness and frequenting public houses as not an aspect of Jewish identity, though acceptable for others. Jews also disliked the loss of control and unacceptable social behaviour that were perceived as consequences of drunkenness, and, unlike Protestants, failed to mention the possibility of alcohol as a method of coping with stress and distress (5). For example in one study, Jewish informants offered a lively  assortment of catastrophic scenarios resulting from drunkenness (6):

“(at a) party at my company just before I joined the firm, a young chap drank too much at the Xmas dinner and told the bosses exactly what he thought of them. It made everybody feel very uncomfortable’;  ‘(drinking leads to) misbehaving socially…blokes and members of the opposite sex tend to be a lot more forward when they are drunk and quite often go over the boundaries of what would be normally acceptable’;  ‘It can lead to immorality – getting into trouble with the police – getting into fights for no reason’; ‘It can lead to abuse or to violence…it can cause husbands hitting wives…destroying furniture, things like this…attacking wives and children’; ‘If you do go to the pub, you expect that there will be a lot of drunk people around and beer flying’.

Those of Protestant background, by contrast, generally suggested rather pleasant consequences from drinking:

‘In small doses…it gives you a buzz, it makes you mingle more’; ‘It helps in social situations…I am more verbally articulate’ ; ‘May help that problem go away altogether’ . ‘If it’s at family occasions where everyone is a bit on edge…once they’ve had a few drinks, everyone is talking…so it does bring down some of the barriers’; ‘People want to enjoy themselves and perhaps put problems to the side’;‘If it makes you feel a little less inhibited and it makes you have some fun then that’s OK’; ‘I find it much easier to talk to people that I don’t know if I’ve had a drink’; ‘I find a drink might relax me’; ‘It drowns your problems’; ‘You lose your inhibitions so you might be able to socialise more’.

We have little evidence about attitudes to drinking in other groups. Muslims have less liberal attitudes to alcoholism than Christians (in Israel) (26). Further evidence from Muslim groups would be of particular interest, given the normative prohibition of alcohol consumption in Muslim society.

GENDER

The alcohol depression-hypothesis arose from the observation that there was gender equality in prevalence of depression in groups in which alcohol consumption is light or negligible. When we compare the prevalence of depression among men and among women in different groups, the evidence indeed suggests that alcohol consumption could be one factor affecting gender differences and similarities in depression – but there are a number of other factors.

There has been longstanding awareness of factors which relate to gender, and which affect depression prevalence. These include differences in lifestyle, in patterns of depressive symptoms, in coping styles including the use of social support, in help-seeking from professionals and others, and in suicide methods and completion rates. But some of the data are consistent with the idea that men and women use alcohol differently, and this affects depression prevalence.

Are differences in alcohol consumption, and associated attitudes, consistent with what the alcohol-depression would predict to be in line with the differences in depression prevalence in these groups? In the general population men consume significantly more alcohol than do women (28). In studies of Jews compared to those of Christian background, Jewish  women and men have been found not to differ significantly in alcohol consumption, or in favourability of attitudes to alcohol, whereas among Christians and those of Christian background, men drink more than women and have more favourable attitudes to alcohol consumption (5, 6, 26).

In Muslim groups, the evidence relating to gender is less supportive of the alcohol-depression hypothesis. Notably, there have been numerous reports that depression among Muslim women is more prevalent than among men, even though alcohol use is negligible In Muslim countries governed by Muslim law, and among religious Muslims elsewhere (15, 16, 17, 27).

INCONSISTENCIES AND GAPS IN THE LITERATURE

There are some inconsistencies and gaps in the literature. 

For example – as mentioned - the literature on depression and gender in Muslim countries is certainly not consistent in showing similar depression prevalence among men and women, as the alcohol-depression hypothesis would predict (15, 16, 17). 

A further undeveloped area relates to the effects of religiosity on both depression and alcohol use among Christians. Religiosity has been assessed using a range of measures, but however assessed, it has been shown to relate to lower consumption of alcohol, and to less liberal attitudes towards alcohol (5, 6, 29, 30). Religiosity has also been shown to have a generally consistent (though weak) relationship with better mental health (31, 32).. These two effects appear to be in conflict with what we might expect from the alcohol-depression hypothesis. If the consumption of alcohol among highly religious Christians is light then we might expect a raised prevalence of depression, but in fact depression prevalence is somewhat lowered among religiously active Christians. There is limited evidence that there is gender equality in depression prevalence among highly religious Christians (12), which is what we might expect from the alcohol depression hypothesis. Nevertheless the evidence indicating an association between better mental health and moderate alcohol use are based on samples of Christian affiliation or background (9).

It is outside the scope of this chapter to systematically consider the influence of  factors other than alcohol use, co-varying with culture and gender, which might influence the prevalence of depression, but we must obviously bear in mind that such factors may be more influential than that of alcohol.

CONCLUSION

The alcohol-depression hypothesis suggests that up to the point of moderate use, alcohol use masks or inhibits depression. It also implies that cultures which support alcohol use will also support beliefs about the benefits of alcohol use. Conversely, alcohol use will be seen as dangerous and bad in cultures which do not support alcohol use. The hypothesis arose from the observation that there is gender similarity in the prevalence of depression in some groups with low alcohol consumption.

Evidence from Jewish groups broadly supports these features of the alcohol-depression hypothesis. Evidence from Christian is incomplete but broadly consistent with the alcohol-depression hypothesis, offering some support for the effects of moderate self-medication with alcohol in lowering depression. Evidence from Muslim groups is generally not consistent with the hypothesis. All the evidence – particularly from Muslims - indicates that that there are many factors other than alcohol use needed to explain the prevalence of depression. Nevertheless there is sufficient to suggest that patterns of alcohol use may be an important contributory factor in depressed mood and depressive disorders.