Positional role changes and drinking patterns: results of a longitudinal study.
Drinking patterns vary throughout the life span (Skog and Duckert, 1993) and these intra-individual variations in drinking patterns have been referred to as the drinking career. Theoretically, it has been argued that drinking patterns can be partly structured by positional roles--e.g., employment, marital and parental--that an individual fills (Knibbe, Drop, and Muytjens, 1987; Neve, Lemmens, and Drop, 1997). Since positional roles change over the life span, one can speculate that they can explain the changes in drinking patterns observed in the course of a lifetime.
For women this theory has been tested in regard to marital, parental and professional roles. For men, however, it has mainly been tested in regard to employment roles, with little consideration of marital and parental roles, reflecting traditional assumptions about gender role division (Gerson, 1993). Nevertheless, the manner in which positional roles are structured and acted has been deeply transformed in the last few years. The now obvious "revolution in women's employment and family patterns" (Gerson, 1993) has modified not only women's roles but also the manner in which men have come to involve themselves in family life and fulfill the responsibilities of fatherhood. In addition, the transformation of the labor market and, more recently, Quebec's high unemployment rate have all the more affected men's pattern of commitment by changing their connection to employment. It has thus become increasingly important to place as much emphasis on how men's drinking patterns are structured by marital and parental roles as on the effect of employment roles. This paper examines how both family and professional roles structure men's drinking careers.
In regard to employment, the results of previous studies are quite debatable. Most studies show that unemployed men drink more frequently and are heavier drinkers than the employed (Crawford et al., 1987; Johnson, 1982; Temple et al., 1991). In addition, Crawford et al. (1987) found that Scottish and English unemployed men, compared with employed men, drink significantly greater amounts of alcohol on their heaviest drinking day, report significantly longer drinking occasions and faster consumption rates, are more likely to report criticism from friends or relatives, and are more likely to report early morning drinking or restlessness/ irritability without a drink. However, based on 10 longitudinal studies from the U.S., Europe and Canada, Temple et al. (1991) reported that chronic unemployment is associated with increased drinking only among young men; among those age 40 or older, it is associated with decreased consumption. On the other hand, in a longitudinal study conducted among 1,980 individuals in the Dutch province of Limburgh, Neve, Lemmens, and Drop (1997) found, on the contrary, that being employed is associated with increased drinking. Their study showed that employed men report higher consumption and frequency of heavy drinking than men who do not hold a job.
Studies on the effects of marriage have clear and homogeneous results: being married is negatively associated with men's drinking. Married men tend to report lower typical quantity per occasion (Celentano and McQueen, 1984; Temple et al., 1991) and less heavy drinking (Neve, Lemmens and Drop, 1997; Parker et al., 1980; Room, 1996; Temple et al., 1991). Conversely, U.S. studies showed that men who are divorced or separated are relatively less likely to abstain from alcohol and relatively more likely to drink at heavier levels (followed closely by the group of single persons) (Celentano and McQueen, 1984; Johnson 1982; Wilsnack, Wilsnack, and Klassen, 1984). Temple et al. (1991) showed that it was specifically for younger men that not getting married and becoming unmarried was associated with increased typical quantity per occasion.
Based on a longitudinal study of young Americans, Horwitz and White (1991) also found that married individuals were less likely to be problem drinkers, but they hypothesized that it might be due to a selection bias by which problem drinkers are less likely to get married. In addition, some studies have suggested that the quality of a given marital relationship may be a better predictor of alcohol consumption than the marital status itself (Horwitz and White, 1991; Richman, Rospenda, and Kelley, 1995).
Very little research on the relationship between men's parenthood and their drinking has been conducted, and consequently no major conclusion about this association has been drawn. On the one hand, Horwitz and White (1991) have argued that the presence of a child may increase men's problematic use of alcohol--not solely because of the presence of the child, but because children may affect the quality of the marital relationship by increasing the occurrence of marital conflicts. On the other hand, Neve, Lemmens, and Drop (1997) reported that having children at home does not affect men's consumption and frequency of heavy drinking. This result is in agreement with Room's hypothesis (1996) that the birth of children in a family may have less effect in changing the drinking of men than that of women. However, it should be noted that so far, even studies concerned with the effects of childrearing on women's drinking patterns have revealed only scattered and inconsistent relationships with heavy drinking (Celentano and McQueen, 1984; Wilsnack and Cheloha, 1987).
Men's multiple roles
While results concerning the effects of marital status are uniform, those related to employment status are less so, and few data are available on the effect of children on men's drinking patterns. However, no research other than that by Neve, Letomens and Drop (1997) has focused on these three positional roles in the same study. Given that men's lives have been through profound changes in recent years, and that culturally it is as valued for a man to be a husband and a father as a worker, it is now imperative that studies focus simultaneously on each type of positional role a man can fill. Because work is no longer the single predominant road to manhood (Gerson, 1993), the spotlight has to be equally on men's family roles and their professional roles. However, it is difficult to evaluate the proportion of men who have actually abandoned work as the predominant road to manhood.
In other respects, studies on positional roles have mainly been transversal, since very few have been concerned with the positional role change issue. One can more precisely understand this phenomenon by conducting longitudinal studies (Room, 1996; Temple et al., 1991) to capture the dynamic processes of these roles. Over a certain period of time, parental and employment roles may shift as individuals get married or divorced, remain married or single, and so forth. Moreover, longitudinal studies allow evacuation of the selection-effect hypothesis (Horwitz and White, 1991) and thereby permit verification of whether positional roles truly influence drinking patterns. To our knowledge, only three longitudinal studies have been concerned with this instability/stability process, and results are not consistent. On the one hand, it has been reported that role changes contribute to a relatively modest amount of the explained variance in drinking patterns (Temple et al., 1991). Horwitz and White (1991), in regard only to getting married versus staying single, obtained the same type of result, concluding that the quality of a role produced well-being or disorder to a greater extent than simply holding a given role. On the other hand, in a study confined to young American women, Hanna, Faden, and Harford (1993) reported that the instability created by a change in social position leads to changes in the drinking pattern, whereas constant status implies only small variations in alcohol consumption.
In view of the above, in order to elucidate the structuring effects of both familial and professional roles for men, the following question should be asked: Do men's changes of positional roles involve changes in drinking patterns? However, in a longitudinal perspective we cannot ignore that, on average, heavy drinkers decrease their alcohol consumption, whereas light drinkers increase it (Skog, 1991). In addition, the well-established fact that people tend to reduce their drinking as they get older also must be taken into account. Consequently, the question can be reformulated: Over and above the effects of baseline consumption and age, do men's changes of positional roles involve changes in drinking patterns?
The data are a subsample of the Quebec Health and Social Survey 1992/93 (QHSS, 1992) and a followup study conducted in 1995 on a subsample of heavy drinkers. The QHSS 1992 is a general population survey representative of the Quebec population (except the northern regions and aboriginal people's territories). Information was gathered on each household member by a face-to-face interview with a key informant in the household and by self-administered questionnaires completed by each household member age 15 or older (n = 22,389). The following analyses are based on the self-administered questionnaire, which, among other matters, addressed drinking behaviors (for more details on measures and sampling procedures, see Audet, 1996).
The 1995 survey was conducted on a subsample of heavy drinkers from the first survey, selected according to their frequency of taking five drinks or more on one occasion. Among the 22,389 respondents to QHSS 1992, 17,194 reported having consumed alcohol at least once over the previous year. The highest quintile (n = 3,353) for the frequency of five drinks or more per occasion constituted the target sample; it corresponds to at least eight occasions during the 12-month reported period. Sante Quebec initiated mail contact with members of that sample in May 1995 (two recalls) to request their participation in the followup study. Of the 3,353 selected respondents, 35% (n = 1,197) agreed to participate in the followup study, 1% refused, 43% did not answer, and 21% of the solicitations came back with the inscription "unknown at this address." It is noteworthy that no followup was done on the target population between 1992 and 1995, and that residential mobility is quite important in Quebec (for example, within the Montreal urban area, residential mobility reaches 20% of the population per year). This likely explains why a large proportion of the target sample was not reached. A mail questionnaire was sent to the 1,197 agreeing participants, and 87.5% (n = 998) returned it completed. The final sample was compared with the target sample, as reported in 1992, according to gender, age, income, education, and drinking patterns. Except for income and education, no significant differences were observed. As with previous findings from self-administered questionnaires, less-educated individuals are underrepresented, creating a self-selection bias toward highly educated individuals (Bourque and Fielder, 1995).
The final 1995 sample consists of 255 women and 743 men. Only men ages 18-60 were considered in our study (n = 617). Men over 60 were excluded, since even though they certainly can change both marital and employment status, it is mostly because they retire or become widowers, and we did not have enough of these specific cases. Data were weighted to correct for the non-proportionality of the sample and the nonresponse rate (using the SPSS weighting procedure).
This study examines changes between 1992 and 1995 on two dimensions of the drinking pattern: the drinking frequency and the frequency of five drinks or more on one occasion. In 1992 the frequency of drinking was an ordinal variable derived from the following question: During the last 12 months, what has been the frequency of your alcohol use? Every day; from 4 to 6 times a week; from 2 to 3 times a week; once a week; from 1 to 3 times a month; less than once a month. The frequency of five or more was a continuous variable assessing the number of times this event occurred over the previous year. In the followup survey, both frequency of drinking and frequency of five drinks or more per occasion were continuous variables, assessing the number of times the event occurred in the previous three months. To estimate the difference between 1992 and 1995 drinking patterns, all four variables were transformed to a 12-month basis.
Changes in three positional roles were examined: marital, parental, and employment roles. The change in marital role is a four-category variable: (1) married at both measurement points (reference category); (2) single at both measurement points; (3) single in 1992 and married in 1995; (4) married in 1992 and single in 1995. Since the last category yielded only 14 respondents, they were excluded from the analysis. The status of married was defined to include common-law union.
Two types of change in parental role were examined: change in parental role and the birth of a child. The change in parental role is a four-category variable: (1) parent at both measurement points (reference category); (2) not parent at both measurement points; (3) parent in 1992, not parent in 1995; (4) not parent in 1992, parent in 1995. To be considered a parent, a respondent had to be living with a child under 18 years of age, whether or not that child was the respondent's. A child's birth is a dummy variable (child's birth = 1). It should be borne in mind that respondents who were parents in both 1992 and 1995, as well as those who were not parents in 1992 but were in 1995, could report a child's birth over the three-year period.
The change in employment role is a four-category variable: (1) employed at both measurement points (reference category); (2) unemployed at both measurement points; (3) unemployed in 1992 and employed in 1995; (4) employed in 1992 and unemployed in 1995. However, in order to perform statistical analysis, persons who were unemployed, were students, or were homemakers were all classified as unemployed, given that 67% of the respondents were employed at both times.
Age and baseline consumption are the two controlled variables used in this study, both being continuous variables. Age ranges from 18 to 60 years old, the mean being 34 years. The baseline annual frequency of drinking ranges from 12 to 365 days. The baseline annual frequency of five or more drinks ranges from 8 to 96 times a year. The measures' distributions are described in Table 1.
TABLE 1 Positional role distribution and, for 1992, drinking patterns and age distribution, means and standard deviations
Men's Final Positional Roles 1992 - 1995 N % Marital status 582 100 Spouse - Spouse 341 58.3 No spouse - No spouse 169 29.1 No spouse - Spouse 72 12.4 Parental status 556 95.5 Parent - Parent 201 34.6 Not a parent - Not a parent 284 48.8 Not a parent - Parent 41 7.0 Parent - Not a parent 30 5.2 A child's birth 90 15.2 Employment status 566 97.3 Employed - Employed 389 66.9 Unemployed - Unemployed 66 11.3 Unemployed - Employed 78 13.4 Employed - Unemployed 33 5.9 Drinking patterns and Age, 1992 N Mean (SD) Annual frequency of drinking 571 137.5 (101.9) Annual frequency of 5+ drinks 582 29.3 (26.8) Age 582 33.7 (10.2)
Method of analysis
Analyses were conducted to assess changes between 1992 and 1995 in the annual frequency of drinking and the frequency of 5+. T-tests were performed, using SPSS 7.0 with the mean value procedure.
Standard multiple regressions, using SPSS 7.0, were used to assess the unique effect of each positional role variable on changes in drinking behavior. Analyses were conducted separately for each drinking measure. All analyses were controlled for age and baseline consumption. Each independent variable was evaluated in terms of what it adds to prediction of the dependent variable that is different from the predictability afforded by all other independent variables, following the method ENTER (Tabachnik and Fidell, 1989). Given the correlation between the birth of a child and being a parent, analyses were done in two steps. In the first step, the "a child's birth" variable was not included in the model, while it was in the second step. Given the fact that the "a child's birth" variable did not modify the relationship between parental role variables and drinking patterns, both type of variables were included in the final model, which is presented in this paper. Due to an insufficient number of cases, no interactions between independent variables were tested. Each regression equation was thoroughly evaluated with regard to statistical assumptions. No problem of multi-collinearity was detected, nor any departure from assumptions of normality, linearity and homoscedasticity of residuals. Some measures had missing data, so final analyses were performed on 505 cases for the annual frequency of drinking and on 517 cases for the annual frequency of 5+.
Change in alcohol consumption
As shown in Table 2, the annual frequency of drinking increased significantly between 1992 and 1995 (t = 3.4; p [is less than] 0.001). In 1992, respondents reported drinking a mean of 137 times a year (std.d. = 102), and this frequency increased to 152 (std.d. = 102) in 1995. On the other hand, the annual frequency of having five drinks or more on one occasion showed no statistically significant change over the three-year period of the study (t = -1.5; p [is greater than or equal to] 0.05), with a reported mean frequency of 29 occasions (std.d. = 27) for 1992 and of 27 (std.d. = 30) for 1995.
TABLE 2 One-sample test on the difference of drinking pattern means between 1992 and 1995
Year Mean Standard Deviation Annual frequency 1992 137.32 101.69 of drinking 1995 151.88 102.06 Annual frequency of 1992 29.23 26.85 5 drinks or more 1995 27.37 29.88 per occasion Test Value = Drinking pattern's mean in 1992 Mean difference t N Annual frequency 14.57 3.390(***) 585 of drinking Annual frequency of -1.86 -1.478 595 5 drinks or more per occasion
(***) p [is less than] 0.001.
Stability at the aggregate level does not necessarily reflect stability in the behavior of individuals. It may reflect that the majority of respondents are stable, or that those who drank more and those who drank less cancel each other out in the aggregated results. In fact, as shown in Figure 1 and Figure 2, important changes in individual drinking occurred between 1992 and 1995.
[Figures 1-2 ILLUSTRATION OMITTED]
Standard multiple regression
A standard multiple regression was performed to examine how changes in individuals' drinking patterns are linked to changes in positional roles. The various drinking pattern dimensions, the [Delta]1992-1995 annual frequency of drinking, and the [Delta]1992-1995 annual frequency of five drinks or more on a single occasion were treated as dependent variables to be predicted by the respondents' changes in positional roles, age, and baseline consumption. Although many variables were not significant predictors of the drinking pattern, they were not withdrawn from the analysis, because they accounted for a certain percentage of the explained variance (Tabachnik and Fidell, 1989). All in all, when age and baseline consumption are controlled, becoming a parent and long-term unemployment have a significant but marginal effect on drinking patterns.
In regard to positional roles, two results were salient. First, the birth of a child between 1992 and 1995 was negatively associated with the [Delta]1992-1995 annual frequency of drinking; respondents who had a child during this three-year period reduced their annual frequency of drinking. Second, to be unemployed at both time points was positively associated with the annual frequency of drinking; this may imply that long-term unemployed persons increased their annual frequency of drinking. No other positional role change was significantly associated with the [Delta]1992-1995 annual frequency of drinking. Not one positional role change was significantly associated with the [Delta]1992-1995 annual frequency of five drinks or more on a single occasion.
Age was significantly--in a very limited way--positively associated with frequency of drinking. As people aged, they tended to increase their annual frequency of drinking. Consequent to the fact that this study was based on a subsample of heavy drinkers, baseline consumption was negatively associated with the [Delta]1992-1995 annual frequency of drinking and with the [Delta]1992-1995 annual frequency of five drinks or more on a single occasion. The higher the consumption level in 1992, the more consumption decreased between 1992 and 1995.
At the aggregate level, we observed significant reductions from 1992 to 1995 in the annual frequency of drinking. On the other hand, we observed relative stability between 1992 and 1995 in the annual frequency of five drinks or more on a single occasion. This result is coherent with what has been observed in Quebec's general population, where the annual frequency of drinking has significantly decreased but the annual frequency of 5+ has not significantly changed between 1989 and 1994 (Kuzminski & Demers, 1998). At the individual level, considerable variation was observed: some men increased their consumption, some reduced it, and some had stable drinking patterns. However, the purpose of this research was to verify whether, over and above the effects of baseline consumption and age, intra-individual variations in drinking patterns over a three-year period could be explained by positional roles--i.e., if changes in positional roles involved changes in drinking patterns.
In regard to the [Delta]1992-1995 annual frequency of drinking, two types of positional role change have a structuring effect: long-term unemployment, and a child's birth. These days, men are more and more what Gerson (1993) calls "involved men"--men who irrespective of their involvement in work are highly committed to parenthood. Now, as with any other type of commitment, parenthood requires commitment in consistent lines of activity. Accordingly, men may have decreased their alcohol consumption because most of the responsibilities associated with a newborn do not mix well with alcohol. Furthermore, although men are increasingly "involved," by no means does this imply that they do not still remain committed to work. As a matter of fact, a result of this study is that long-term unemployment is associated with an increased annual frequency of drinking. We might hypothesize that this result is due to the fact that for men work is still the predominant road to an everyday structured life, and that in agreement with Knibbe, Drop and Muytjens' results (1987), men with a less structured everyday life are those with a higher chance of intensified consumption.
Another basic finding of the analysis is that no other positional role changes affected significantly the [Delta]1992-1995 annual frequency of drinking and that not a single one affected significantly the [Delta]1992-1995 annual frequency of five drinks or more on a single occasion. A first look at this study's results may be surprising. Although the way marital, parental and employment roles structure drinking patterns was not consistent from one study to another, nearly every study mentioned in this paper reported that positional roles do in fact structure drinking patterns. However, we cannot ignore the few studies that showed that roles do not play a major role in drinking patterns. Moreover, Temple et al. (1991) emphasized that although positional role changes can explain certain tendencies in regard to drinking patterns, many of those tendencies are not significant. In fact, those authors concluded that positional roles contribute to a relatively modest amount of the explained variance in drinking patterns. Our results are similar, in that positional role changes do not explain a large part of drinking behaviors.
In general, explanations of those null findings may concern the qualitative aspects of the positional role rather than the role in itself. A first explanation may be that positional roles structure drinking patterns because of what they imply in terms of responsibilities. All three positional roles under study in this paper imply patterns of commitment, although different ones. First, employment may be a commitment due to the set of regular activities that it implies. Second, marriage may also be a commitment but (hopefully) a less constraining one, since it is due not to regular activities, but to attachment. Third, parenthood may be the most demanding commitment, since it refers both to regular activities (feeding, changing diapers, bathing, etc.) and to attachment. Hence it may not be different positional roles per se, but different responsibilities (which may vary for the same positional role) that determine drinking patterns. However, our data did not permit us to test such a hypothesis; we would have needed data on both positional roles and responsibilities.
A second explanation may be that, as some authors (Horwitz and White, 1991; Richman, Rospenda and Kelley, 1995) have reported, the quality of a role relationship--the nature of marriage, parenthood, work, etc.--may explain drinking patterns to a greater extent than does the mere holding of those positional roles.
A third hypothesis may be that overall drinking careers are stochastic processes. Skog and Duckert (1993) pointed out that individuals' life histories are influenced by numerous life events of different types, which are very complex; consequently there is no reason why drinking careers should be any different. Given that, it may also be possible that the context in which positional role changes occur, and the individual vulnerability to such events, contribute to more explained variance among drinking behaviors than do changes of positional roles as such (Gorman and Brown, 1992; O'Doherty and Davies, 1987). Neve (1997) has proposed that a stress and coping framework should be incorporated in positional roles analysis in order to increase its explanatory power. Nevertheless this is very hard to assess in a population survey design.
Finally, this study was conducted with limitations that may have affected the results. First of all, the sample size was not large enough for examining interactions. It was not possible to follow Temple et al. (1991) in showing interactions between positional role changes and age, nor was it possible to verify interactions between different positional role changes. The latter type of interaction may have had the most salient effect, given the possibility that positional roles have cumulative rather than single effects. In fact, studies among female populations have shown that it is role conflict that explains drinking behaviors (Johnson, 1982), while others found that explanation lies rather with role deprivation (Parker et al., 1980; Shore and Batt, 1991; Wilsnack and Cheloha, 1987; Wilsnack et al., 1984). Due to an insufficient number of cases, we were not able to test either of those theories.
Another limitation of this study was that in regard to positional roles, respondents were relatively stable: 58% of the respondents were married at both study periods, 88% did not change parental status, and 67% were employed at both times. Given that stability, some role categories had to be collapsed to be able to perform statistical analysis. For example, those who reported being unemployed, students, or homemakers were all included in the "unemployed" category, so information may have been lost.
The present research suggests that it may not be positional roles as such that determine drinking patterns. On one hand, this result may be explained by the methodological limitations of the present study. On the other hand, it may be because it is not positional roles in themselves that influence drinking patterns but rather what is inherent in these roles, such as their quality and the stress or degree of commitment attached to them. Advancing understanding of how changes in drinking behaviors occur should therefore consider the qualitative aspects of positional roles. Such research may well require complementary quantitative and qualitative research.
TABLE 3 Regression of drinking measures on changes in positional roles, age, and baseline consumption
[Delta] annual [Delta] annual frequency frequency of drinking of 5 drinks or more per occasion Positional roles 92-95 Unstandardized Coefficients Constant 29.7 11.51(*) Marital Spouse - Spouse (reference) (reference) No spouse - No spouse -4.7 -1.6 No spouse - Spouse -9.5 -2.8 Parental Parent - Parent (reference) (reference) Not a parent - Not a -7.1 2.7 parent Not a parent - Parent -6.3 -3.1 Parent - Not a parent -17.5 3.0 No child's birth (reference) (reference) A child's birth -26.7(**) -4.6 Employment Employed - Employed (reference) (reference) Unemployed - Unemployed 31.8(**) 4.7 Unemployed - Employed 11.2 4.9 Employed - Unemployed 6.8 3.5 [R.sup.2] Positional 0.02 0.003 roles Age 1.0(*) -0.1 Baseline consumption -0.3(***) -0.4(***) Adj. [R.sup.2] 0.15 0.169 Total adj. [R.sup.2] 0.17 0.17 Number of valid cases 505 517
(*) p [is less than] 0.05.
(**) p [is less than] 0.01.
(***) p [is less than] 0.001.
AUTHORS' NOTE: Acknowledgments go to Michael McCubbin for having reviewed and made substantial comments to this paper. An earlier version of this paper was presented at the 23rd annual Alcohol Epidemiology Symposium, Reykjavik, Iceland, June 3, 1997.
Audet, N. (1996). Cahier technique and methodologique; Enquete sociale et de sante 1992-1993. Vol. 2: Definition et composition des indices et regroupements. Montreal: Sante Quebec.
Bourque, L.B. & Fielder, E.P. (1995). How to conduct self-administered and mail surveys. The survey kit, vol. 3. Beverly Hills, CA: Sage Publications.
Celentano, D.D. & McQueen, D.V. (1984). Alcohol consumption patterns among women in Baltimore. Journal of Studies on Alcohol, 45(4): 355-358.
Crawford, A., Plant, M.A., Kreitman, N. & Latchman, R.W. (1987). Unemployment and drinking behavior: Some data from a general Population survey of alcohol use. British Journal of Addiction, 82: 1007-1016.
Gerson, K. (1993). No man's land: Men's changing commitments to family and work. New York: Basic Books.
Gorman, D.M. & Brown, G.W. (1992). Recent developments in life-event research and their relevance for the study of addictions. British Journal of Addiction, 87: 837-849.
Hanna, E.Z., Faden, V.B. & Harford, T.C. (1993). Marriage: Does it protect young women from alcoholism? Journal of Substance Abuse, 5(1): 1-14.
Horwitz, A.V. & White, H.R. (1991). Becoming married, depression, and alcohol problems among young adults. Journal of Health and Social Behavior, 32 (September): 221-237.
Johnson, P.B. (1982). Sex differences, women's roles and alcohol use: Preliminary national data. Journal of Social Issues, 38(2): 93-116.
Knibbe, R.A., Drop, M.J. & Muytjens, A. (1987). Correlates of stages in the progression from everyday drinking to problem drinking. Social Science Medicine, 24(5): 463-473.
Kuzminski F. & Demers, A. (1998). Evolution (1978-1994) et determinants de la consommation d'alcool et de drogues au Quebec. (Analyses secondaires des donnees d'enquetes nationales). Rapport final. (Quebec: MSSS).
Neve, R. (1997). Changes in alcohol use and in drinking problems in relation to role transitions in different stages of the life course. Paper presented at the 23rd annual Alcohol Epidemiology Symposium of the Kettil Bruun Society for Social and Epidemiological Research on Alcohol, Reykjavik, Iceland.
Neve, R., Lemmens, P. & Drop, M.J. (1997). Gender differences in alcohol use and alcohol problems: Mediation by social roles and gender-role attitudes. Substance Use and Misuse, 32: 1439-1459.
O'Doherty, F. & Davies, J.B. (1987). Life events and addiction: A critical review. British Journal of Addiction, 82: 127-137.
Parker, D.A. Parker, E.S., Wolz, M.W. & Harford, T.C. (1980). Sex roles and alcohol consumption: A research note. Journal of Health and Social Behavior, 21 (March): 43-48.
Richman, J.A., Rospenda, K.M. & Kelley, M.A. (1995). Gender roles and alcohol abuse across the transition to parenthood. Journal of Studies on Alcohol, 56(5): 553-557.
Room, R. (1996). Gender roles and interactions in drinking and drug use. Journal of Substance Abuse, 8(2): 227-239.
Shore, E.R. & Batt, S. (1991). Contextual factors related to the drinking behaviors of American business and professional women. British Journal of Addiction, 86:171-176.
Skog, O.-J. (1991). Implications of the distribution theory for drinking and alcoholism. In: Pittman, D. & White, H.R. (eds.), Society, culture and drinking patterns reexamined, 576-597. New Brunswick, NJ: Rutgers Center for Alcohol Studies.
Skog, O.-J. & Duekert, F. (1993). The development of alcoholics' and heavy drinkers' consumption: A longitudinal study. Journal of Studies on Alcohol, 54(2): 178-188.
Tabachnik, B.G. & Fidell, L.S. (1989). Using multivariate statistics (2nd ed.) New York: HarperCollins.
Temple, M.T., Fillmore, K.M., Hartka, E., Johnstone, B.M., Leino, E.V. & Motoyoshi, M. (1991). A meta-analysis of change in marital and employment status as predictors of alcohol consumption on a typical occasion. British Journal of Addiction, 86(10): 1269-1281.
Wilsnack, R.A. & Cheloha, R. (1987). Women's roles and problem drinking across the lifespan. Social Problems, 34:231-248.
Wilsnack, R.W., Wilsnack, S.C. & Klassen, A.D. (1984). Women's drinking and drinking problems: Patterns from a 1981 national survey. American Journal of Public Health, 74(11): 1231-1238.
CATHERINE PARADIS worked on this study as a graduate student at the Health and Prevention Social Research Group (GRASP) at the University of Montreal (2801 Edouard-Montpetit, Montreal, Quebec H3C 3J7, Canada; firstname.lastname@example.org) and is currently engaged in dissertation research at the National Institute for Alcohol and Drug Research in Oslo. ANDREE DEMERS, on the faculty of the University of Montreal, leads an alcohol and drug research group at GRASP and has published a number of papers on the sociology and demography of drinking. LOUISE NADEAU, also on the University of Montreal faculty, has published a textbook on alcohol (Vivre avec l'alcool, 1990), and several analyses of women and drinking, including the book Les femmes et l'alcool en Amerique du Nord et Quebec (1984).3
|Printer friendly Cite/link Email Feedback|
|Author:||Paradis, Catherine; Demers, Andree; Nadeau, Louise|
|Publication:||Contemporary Drug Problems|
|Date:||Mar 22, 1999|
|Previous Article:||The impact of credit and debit cards in liquor stores on opinions and consumption patterns.|
|Next Article:||Lesbian social drinking: the role of alcohol in growing up and living as lesbian.|