For research on

Parental Role Model: Abstinence in Best

The Family Transmission of Alcohol Abuse

79 Studies and Counting

by Thomas E. Radecki, M.D.

(The author apologizes that this review has not been updated since the early 1990’s. However, the evidence is so extremely overwhelming that one can be sure that the newer studies will only be a rehash of the older ones.  Indeed, the several studies which I have seen covered in the popular media all agree with the conclusions below.)

Introduction

The impact of an abstinence message for alcohol has been debated by humans for at least 2400 years and by substance abuse researchers since at least 1940. Unfortunately, scientific data looking at this issue is only recently available. Some of the early academic studies were flawed with conclusions drawn far out of line with what their own figures showed. Some of these early researchers attacked the concept of teaching abstinence, claiming that such a policy causes more problem drinking, not less. Examples are Skolnick, (1958), Globetti, (1967, 1978, 1991), and Linsky (1986). In actuality, in almost every case their own data supports the benefits of teaching abstinence as is described later in this paper. There is not a single scientifically valid study that shows that teaching teenagers to drink is as wise as teaching them to abstain.

In addition to misinterpreting their own findings, the early researchers noted above never looked at parental drinking behavior and were thus much less precise than more recent work. Sometimes, studies that clearly show that abstaining parents have the fewest children with alcohol problems (e.g. Hamburg, 1982) are cited with the claim that they show that abstaining by parents conflicts with community and peer norms and adversely affects their offspring's alcohol use (Friedman, 1985).

Of 79 studies which look at actual parental drinking patterns in non-alcoholic parents, 76 studies find at least a harmful tendency for parental alcohol drinking to increase drinking in the offspring. Although much smaller in number, all but one study reporting abstaining, light, and moderate drinking parents separately have found that the less parents drink, the less likely that their children will be heavy and problem drinkers. The only partial exception to this is a master's thesis with a very small student sample (Fillmore, 1970) which found several heavy drinkers among the offspring of abstaining parents. This may have been a chance finding due to the very small sample size. Another interesting report found a higher level of alcoholism in the family histories of abstainers, which, if replicated, suggests that offspring of abstaining parents are doing better despite a possibly somewhat greater genetic loading for alcoholism. Those studies which have looked at actual problem drinking in offspring all find fewer problems in the children of non-drinking parents than for the children of light or moderate drinking parents.

The studies of the transmission of alcohol abuse by non-alcoholic drinking parents do yield important information for parents and professionals. They have repeatedly shown that parental role models do play an important role in predicting later alcohol abuse. They show a non-drinking role-model is best.

A huge amount of research shows that peer drinking and attitudes are also very important, but none of the research examines how much parents influence the selection of peers and how much of the peer drinking is an indirect impact of parental modeling from the parents of the peers, parental attitudes, and lack of parental limit setting. Extrapolating from the research, we can say that other important role models, e.g. relatives, friends, teachers, church, television, music and movie heroes, and local and national leaders have or can have important effects as well.

Besides cross-sectional studies of Americans, the parental drinking effect has also been found in Chinese-Americans, African-Americans, Hispanic-Americans, and Korean-Americans. Studies in England, Ireland, Scotland, Canada, Norway, Poland, Sweden, Australia, and New Zealand have also noted that non-drinking or very light drinking by parents is linked with their children having lower rates of heavy drinking or fewer negative outcomes. The Swedish study even found a lower death rate in the children on non-drinking parents! Two small studies with Philippino-Americans and Japanese-Americans found only weak, non-significant trends of parental drinking influencing heavier drinking in their children. As noted above, no study has found that light to moderate drinking by parents results in fewer alcohol problems in their offspring compared to parental abstinence.

Non-drinking parents have fewer children who later have alcohol abuse problems than light, moderate or heavy drinking parents. Most studies of children of alcoholics also find a higher rate of drinking problems in the children of alcoholics when compared to non-drinking and light-drinking parents. Although very light drinking parents (drinking less than monthly) have been separated out from light drinking parents in only a couple studies, each of those studies did show a small trend for the offspring of non-drinking parents to do still better. The conclusion from the research is that parental abstinence is best, and a very light drinking parental role model is second best.

Additional research reviewed below shows that religions that preach against alcohol have fewer followers who end up as heavy drinkers. One study even demonstrated that followers of other religions in the same region as a dominant abstinence oriented religion also drink less than followers of those same other religions in other parts of the country. These studies again suggest that an abstinence message is helpful, not harmful, even for those who end up drinking.

Another line of research reviewed below, i.e. the impact of the parental message on underage drinking shows that a firm "no drinking" message from parents has a better effect than a weak message or one condoning underage drinking. Yet another line of research reviewed below shows that parental modeling of other drug use, i.e. nicotine or illicit drugs, has a similar effect of increasing use of those and other drugs.

Studies on Non-Alcoholic Parental Drinking Role Models

1) Probably the best study on the effects of parental role-modeling is that of Grace Barnes and John Welte (1990). Interviews with 6,364 randomly selected New York young adults found abstaining by fathers and abstaining by mothers each had a positive effect. Each was linked to less heavy drinking and fewer alcohol-related problems in the adult lives of their children. This effect was true for both sons and daughters. The study found that in families where parents did drink, the more alcohol consumed, the more frequent were drinking problems in the lives of the children in their own adult life. Growing up in a family where the father was absent had an additional negative effect over and above the drinking pattern of the mother. Also, the earlier the first drink and the first time drunk, the more likely that drinking problems later developed.

2) A much smaller but prospective study by Harburg starting in 1960 in Tecumseh, Michigan, a rural town with 10,000 citizens, followed 420 family sets of a father, mother, and son or daughter for 17 years. The study found that parental drinking patterns do have an important effect with the strongest effect coming from a non-drinking father. Abstaining same-sex parents had the smallest percentage of children who grew up to be heavy drinkers. These abstaining parents were followed stepwise by parents who were very light drinkers, light drinkers, moderate drinkers and lastly heavy drinkers. The only exception to this progression occurred in children of very heavy drinkers who tended to have fewer drinking problems, as if reacting against the aversive role model that they had grown up with.

In this study, the drinking pattern of the same sex parent had the strongest effect. Families where both parents abstained had by far the fewest sons who were heavy drinkers, fewest who were moderate drinkers, and by far the most who were abstainers. The amount of alcohol consumed per week by same-sex offspring increased in a stepwise fashion from of offspring of abstaining parents (2 oz.), to of very light drinkers (3 oz.), light drinkers (6 oz.), moderate drinkers (7 oz.) and heavy drinkers (9 oz). The alcohol consumption by the same-sex offspring of the small number of very heavy drinkers declined (5 oz). (Figure 1.)(Harburg, 1982, Webster, 1989).

Harburg's and Barnes' figures once again strongly disprove the notion that abstaining by parents causes both a higher percentage of abstainers with a higher percentage of offspring who become heavy drinkers. While the former is true, the later is absolutely not true.

Other studies are listed below by year of publication:

3) Straus (1953) reported a national sample of over 15,000 college and university students from 27 schools. Student drinking patterns were strongly related to the drinking patterns of their parents. Among men, 92% of the students who came from homes in which both parents drank were themselves drinkers, compared with only 58% of those from homes where both parents abstained. Among women, the respective percentages were 83% and 34%.

4) Haer (1955) conducted a Washington state-wide cross-sectional survey and found that the drinking of 478 adults was highly correlated and roughly equally with their recall of their friends', spouses', fathers' drinking, and mothers' drinking. It may be that people tended to pick out friends and spouses who had similar drinking patterns, but the study demonstrates that parental drinking patterns do have a definite and lasting impact on the drinking behavior of their children. Again, abstaining parents had the smallest percentage of frequent drinking offspring (three or more times/week) and very light, light-moderate, and very frequent drinkers had an increasingly larger percentage of their children grow up to be very frequent drinkers. Parental impact was strong (p <.001) for both males and females. The percentage of abstainers in the study was quite high compared to modern times (i.e. 35%), but very low compared to their parents generation (65% abstainers) demonstrating the tremendous increase in alcohol consumption in the U.S. since the late 1920's and especially since the repeal of prohibition.

5) Gusfield (1961) studied 185 Eastern US undergrad males. High users of beer were more likely than low users to have parents who were high users of beer. Students of these parents were more likely to join fraternities and fraternity members in the study were almost twice as likely to be heavy drinkers.

Irme (1963)

Prove (1965) found an association between alcohol consumption of parents and their children.

6) A Polish study found that drinking by children increased with the frequency of parental drinking and with the drinking of vodka by the father in the week preceding the survey (Leowski, 1968). Drinking was fairly common among the 495 11- to 15-year-olds studied.

Akers (1968)

7) Gross patterns of adolescent drinking described across several studies associate the child's with the parent's drinking style (Bacon, 1968).

8) Riester and Zucker (1968) found parental alcohol use associated with teenagers' alcohol use in a mid-Atlantic community. There were many nonusers of alcohol as well as heavy users among the children of heavy-use parents.

9) Wieser (1968) studied 408 Germans 14 and older. There was a high correlation between drinking patterns of respondents and their parents (p = .54). Weiser didn't asked about drinking problems. 32% of parents were indifferent or permissive with respect to children's drinking, 61% were against it, 56% were permissive of drinking in general, 38% were against. Germans have a high rate of drinking and alcohol problems.

10) Cahalan (1969) indicated that heavy drinking among both men and women was positively associated with the respondent's recalled frequency of parental drinking.

11) Fillmore (1970) reported that children of two heavy drinkers tended to be either heavy drinkers or abstainers themselves. Her master's thesis study was only 150 high school students asking them about their own and their parents' drinking. When both parents abstain, children are more likely to abstain, although she found several heavy drinking offspring of abstainers.

12) Forslund (1970) found peer and parental alcohol use strongly correlated with self-reported use of alcohol although father's drinking had a significant effect only on daughters. The study was of Albuquerque high school students.

Lawrence (1970)

Stacey (1970)

Jones (1971)

13) Edwards' random-selection interview study (1972) of 928 suburban London adults (89% of those sampled) found that the quantity of sons' and daughters' drinking increased stepwise with fathers' drinking. Abstaining or very light drinking fathers had the most abstaining sons and daughters and fewer moderate or heavy drinking offspring. Mothers' drinking was somewhat less influential for their sons although still significant and very much related to how their daughters drank. Abstaining or very light drinking fathers had half as many heavy drinking sons (8%) as light or moderate drinking fathers (15%) with heavy drinking father's still higher (24%). A similar pattern was true for their daughters except that the heaviest level of drinking for daughters was moderate drinking, not heavy drinking. Of males 7% were abstainers and females 10%; 17% of males were very light drinkers and 36% of females (less than monthly). Differences in consumption by social class were minimal. Many more men than women were heavy (15% vs 1%) or moderate (27% vs. 7%). Irish, Scottish and Catholics had twice as the percentage of heavy drinkers as the English and members of the Church of England. The upper classes drank much more often at home of in restaurants while lower classes were much more likely to drink in pubs.

14) In a well done study by Smart (1972) using a questionnaire survey of 8865 Toronto 6th, 7th, 9th, 11th, and 13th grade students and their parents, parental use was correlated with student use of each of 12 drugs measured, including alcohol. The relationship was very strong when parents and children were using the same drug, e.g. of students reporting daily maternal use of tranquilizers, stimulants, and barbiturates, 31, 25, and 21%, respectively, used these drugs vs 4, 5 and 3% of student users whose mothers were nonusers. Mothers' use of alcohol was more closely related to students' use than was fathers' although both were significant.

15) Wallace (1972) found in a study of 3954 Norwegian adults that 79% of men and 62% women drank. Drinking was more common in large communities, where the sale of alcohol was least restricted by law, and where there was a minimum of religious and abstaining influences. Growing urbanization was attributed for the increasing alcohol consumption. Degree of religious involvement, parents' abstinence, and sex best predicted drinking. Heavy and problem drinking were not examined.

16) Adler (1973) administered a questionnaire to 1600 10th to 12th graders in five Pennsylvania mill towns. Parents who were non-drinkers had the most non-drinking offspring and the fewest habitual alcohol users or habitual drug users. Although the difference were small, occasionally and weekly drinking parents' children did not do as well. The children of parent who drank three times a week or more did much worse. Interestingly, non-drinking children were most likely to say that their world was exciting, and not boring, while habitual drug users found their world boring and not okay. Nonusers were most likely to say they were most influenced by their parents, while habitual drug users said they were most influenced by their friends.

Braucht (1973)

Mylander (1973)

17) McKilip (1973) studied sibling, peer and parent use of both licit and illicit drugs in 628 white, middle-class Chicago high school students using a questionnaire. Teen use most strongly correlated with peer drug use followed by sibling and parental use.

18) Annis (1974) studied 539 Canadian adolescent homes, interviewing each parent and teen separately. Multiple drug use was common. He found clear evidence of imitative learning by teens of specific drug use through parental example and the generalization of this learning to the use of other drug categories. When mother and/or father used alcohol, their teen was more likely to drink. Parental cigarette use also increased their teens alcohol use. Lawrence (1974) found mother's influence to be greater.

19) Hanson (1974) parental drinking patterns strongly associated with their children's use of alcohol. Hanson surveyed on 3696 students at 37 colleges and universities spread across the U.S. in 1970-71.

20) Orford (1974) studied 1323 college students in England. A 44-page questionnaire was mailed with a 76% return rate. He found that parents' drinking and approval of drinking were correlated with their offspring's drinking and their attitudes toward drinking. Alcohol related problem behaviors were quite common among the students. Orford did conclude that peer use was still more important.

21) Parfrey (1974) studied 444 Cork, Ireland college students. Non-drinkers were less likely to smoke and none used marijuana although 30% of the heavier drinking category did. Of men 20% and women 36% abstained. Abstaining by parents was even higher, 31% of fathers and 56% of mothers. Only mothers' and daughters' drinking were reported associated although quantity factors were not analyzed. There was a trend for increased drinking by sons of drinking mothers. Non-drinking mothers had slightly fewer heavy drinking sons although the difference was not significant. Data for fathers was not reported. Interestingly, abstaining students were much more likely to want to marry an abstain and drinkers to marry drinkers.Smoking and marijuana use were strongly correlated with drinking and religious attendance inversely correlated. Heavier drinkers started drinking at an earlier age and in bars or hotels. Sibling drinking was strongly associated with college student drinking and heavy drinking.

22) Stenmark (1974) found a high correlation between fathers' alcohol use and the use of alcohol, marijuana, and other drugs by their offspring.

23) Parker (1975) studied 218 female undergrad in the U.S. and found a father's heavy alcohol consumption was related to his daughter's heavy consumption.

24) A Research Triangle Institute reports indicates that 57% of adolescent offspring of abstaining parents were abstainers compared to 15% of those who perceive their parents as drinking on a regular basis. Of teens of non-drinking parents, 24% drank at least moderately vs. 51% of those whose parents drink regularly. Students whose parents drink were more likely to be moderately heavy and heavy drinkers than those whose parents abstain in a study of 15,000 adolescents grades 7-12 (Rachal, 1975). Harford (1976) using the same data base reports blacks had the highest rate of abstaining and the lowest rate of heavy drinking. Likewise, Southern youth had the highest rate of abstaining and the lowest rate of heavy drinking. Non-drinkers were much more likely to have non-drinking peers and vice versa.

25) O'Connor (1976) interviewed 774 Irish, Anglo-Irish, and English 18-21-year-olds and 613 fathers and 747 mothers. Parental attitudes had the stronger influence on children's drinking patterns although the parents' actual drinking did as well. Drinking patterns were rated from abstinent to very heavy. If both parents were heavy drinkers and approved of their children drinking heavily, odds were 3.5 to 1 that the child would be a heavy drinker. If father drank heavily and mother lightly but disapproved of heavy drinking by their children, the odds were 1.5 to 1 their child would be a light drinker. If they approved of heavy drinking by their child, the odds were 3.4 to 1 their child would drink heavily.

Smart (1976)

26) Margulies (1977) in a multiphasic random sample of 8206 public secondary students in New York State did two surveys six months apart. Full data on 1936 teens was analyzed. Most of the teens were already drinking distilled spirits at the beginning of the study. Of the initially non-spirit drinking teens, those closer to peers than parents were most likely to start drinking. Although parental attitude on distilled spirits had some impact on onset of use, the actual example set by parents was more important with father's example being somewhat more important. Number of friends drinking and best friend's attitude about the harm of drinking were also significant. Those dating and attending parties were most likely to start drinking. Cigarette, beer, and wine users were more likely to start drinking spirits.

27) Donovan (1978) surveyed 7,481 7th to 12th graders and found students who were categorized as problem drinkers to have parents who drank significantly more than parents of drinking students who were not categorized as problem drinkers. Non-drinking students were not included in the study.

28) Hawker (1978) studied 7306 British high school students with a questionnaire. Parents approval and drinking was linked to more drinking by their children.

29) Kandel (1978) found that 82% of drinking families raise youth who also drink, while 72% of abstaining families produce abstainers. A baseline and 6 month random sample of New York public secondary schools students with parent and peer questionnaires as well totalled 5,423 students with a 61% parent response rate. Peer and parental influences were synergistic. Survey included cigarettes, alcohol, marijuana, and other illicit drugs.

30) Smart (1978) surveyed 1,439 Ontario high school students. Father's drinking was a significant predictor of offspring intoxication. Friends' drinking accounted for 11% of the variance in drinking behavior, father's drinking explained 5.7%, and mother's drinking 3.7%. No surprisingly, intoxicated drinking is done away from the house without the parents' knowledge.

31) Fillmore (1979) surveyed adolescents and their parents and found the attitudes and behaviors of parents are good predictors of adolescents' drinking habits.

32) Fontane (1979) studied 99 eastern US undergrads. Parental beverage preference and reported intoxication was related to student, especially male student, preference and intoxication.

33) Kaplan (1979) studied 163 western US undergrads and found a history of parental drinking was related to heavy student consumption.

34) Wechsler (1979) surveyed 7083 New England college students in 1977. More than 90% of the students came from households where at least one of the parents drank. 92% of students already drank before getting to college. Drinking increased somewhat for each year in school although the 3.7% rate of abstinence was steady throughout. 30% of men were heavy drinkers but only 11% of women. Whites males (31%) were much more often frequent-heavy drinkers than blacks (16%) or Asians (11%). Catholics were most often heavy drinkers and Jews least often although still 16% of Jewish men and 5% of Jewish women fell in the heavy drinking category. Social class did not play much of a factor. Heavy drinking was 50% less common in students with better academic performance. Of men 71% had been drunk in the past year and 54% of women. 25% of men and 14% of women got drunk at least once a month. Nearly 25% of men reported they had been in trouble with authorities and 20% in physical fights after drinking. About 10% also reported that drinking had caused them to have an automobile accident and a similar proportion said they had been in other alcohol-caused accidents in which someone was hurt or had seriously damaged a friendship as a result of drinking. 15% reported blackouts. In the heavy drinking environment of 1977, only 7% of the children of abstaining fathers ended up being abstainers themselves. This was still higher than for light/moderate (3%) or heavy/problem (2%) drinking fathers. The rates of heavy drinking by the male children of abstaining fathers was high (26%), but not as high as for infrequent drinking fathers (27%), light/moderate fathers (35%) or heavy/problem drinking fathers (39%). Drinking in high school and the level of drinking in high school did predict heavy drinking in college. Abstaining or infrequent/light drinkers were much less likely to be cigarette or cannabis smokers or users of other illicit drugs with a stepwise progression up to frequent/heavy drinking students. Harmful consequences from drinking increased steadily as the quantity drank increased with 1% of male infrequent-light drinkers suffering an alcoholic blackout compared to 33% of frequent heavy drinkers. 44% of frequent-heavy male drinkers had trouble with authorities vs. 4% of infrequent-light drinkers. 19% of frequent-heavy male drinkers had an automobile accident due to alcohol vs. 1% of infrequent-light drinkers.

35) Cutter (1980) studied 128 northern U.S. undergrads. Maternal deviant drinking and weak parental norms about drinking were related to increased student drinking problems.

36) Huba (1980) found peer and parental alcohol use strongly correlated with self-reported use of alcohol in a study of 1,634 Los Angeles 7th to 9th graders. Includes a 6 month follow-up. 13 illicit and licit drugs were covered. Peer influence became more important for older students

37) Lassey and Carlson (1980) reported that parental drinking behavior was strongly associated with that of offspring's among rural youth. Frequent drinking 8th graders and 12th graders were much less likely to have non-drinking parents. Peer behavior was also important.

38) Jaquith (1981) Criminol (Aug). 3000 7th to 12th graders in Midwest studied. Peer influence direct effect on alcohol and marijuana use and indirect through the subject's internal definition of appropriate behavior. Parental influence important for alcohol although less than peers.

39) Reiskin (1981) found no significant relationship between student intoxication and parent drinking practices in 252 New England undergrad students and in an additional 217 students using the mental health service. 95% drank, 63% had been intoxicated in the past year, 62% used cannabis. Drinking was strongly associated with the use of cannabis. Actual figures on parental and youth drinking patterns are not reported.

40) Wechsler (1981) studied 312 heavy drinking New England undergrads and compared them to 6383 lighter drinkers. Parental drinking pattern was reported as not related to heavy-drinking student consumption, although no figures are not given. Heavy drinking was related to a student's more frequent drinking pattern in high school.

41) Clayton (1982) sampled 2,510 men nationally in 1974 aged 20-30 for eight classes of licit and illicit drugs by interviews in 1974-5. Mother's and father's drinking patterns when respondent was 15 were used. Current friends' use followed by wife or partner's use were most important. Sibling use also had an impact. Parental drug behavior minimally associated with drug use and future intentions.

42) Rooney (1982) surveyed 4,941 high school seniors from the northeast in 1977 and found peer drinking the best predictor of youth drinking. Maternal alcohol use was correlated with teen beer use (p <.001), but not paternal use although actual figures are not given. The impact of maternal alcohol use disappeared on multivariate analysis. Beer use by friends was a much stronger factor and were their own standards and the standards of fellow students.

43) Christiansen (1983) found the maternal drinking behavior in his study was a stronger predictor of adolescent drinking than was the paternal drinking behavior. Their study had 1600 youths. They also found problem parental drinking related to adolescent drinking although only weakly.

44) Fawzy (1983) of UCLA studied 262 Los Angeles area 13- to 17-year-olds and their parents. Parental use of addictive substances were moderately to strongly related to use by their children. His may be the only to include coffee in the list of addictive substances studied. Strong relationships were found between parental coffee use, mother's use of cigarettes, and father's use of hard liquor and substance abuse by their children. Fathers were averaged one to two drinks per day had the most offspring with substance abuse, more than fathers who drank an average of five drinks a day, a probably alcoholic pattern. This has 40% of the children were Spanish speaking, although 38% also spoke English. For purposes of the study, a teen user was one who during the month before the initial interview, drank alcohol or used illegal drugs.

Coffee consumption by parents using over 10 cups per day was linked with 82% of their offspring using being substance users. Parents drinking 1-10 cups per day had 60% of their offspring substance users. Parents who abstain from coffee had 37% of their teenage children using substances.

Daily drinking fathers had 72% and daily drinking mothers 77% of offspring substance users. Weekly drinking parents had 57% teen users and parents drinking less than once a month or not at all had only 46% of the teenage children users. Of parent who used marijuana, 80% of their children were substance users. Parental smokers had 75% of their children substance users. Fawzy concludes that "parental abstinence from substances may be the first line of defense against youthful initiation and efforts to maintain a quality relationship with their children may be the best hedge against emotional pathologies."

45) Blount (1984) found peer and parental modeling effects in black adolescent drinking.

46) Byram (1984) did a random sample of 7-12 graders in the South totaling 335 students using self-report questionnaires. Adolescent alcohol use increased as friends' use of alcohol increased, family closeness diminished, and adult use of alcohol increased.

47, 48, 49) Forney (1984, 85, 87) found similar relationship between student drinking and parental drinking behaviors in studies conducted among middle school and high school students. One (1984) looked at 1,715 6th and 8th graders. Parental drinking patterns had an influence over junior high children's drinking patterns. Abstaining fathers were most likely to have abstaining children and least likely to have frequent or heavy drinking children. Frequent or heavy drinking parents were most likely to have frequent of heavy drinking children. She reports her studies suggest that drinking patterns may be set early in life, influenced by parental role models.

50) Mookherjee (1984) found that in families where one or both parents drink, the children are more likely to drink.

51) Rittenhouse (1984) studied NIDA national survey data of parent, sibling, and 14-17-year-old teen tobacco, alcohol and marijuana use. He found mothers' and older siblings' substance abuse correlated with teen substance abuse but that the fathers' substance abuse did not.

52) Dishio (1985) in a study of 136 7th and 10th grade boys found the degree of parental monitoring and enforcement of rules to be negatively correlated with their children's alcohol use. This correlation was stronger that the positive correlation between the parents' alcohol use and the child's alcohol use although this later correlation was lower than in most other studies.

53) McLaughlin (1985) found parental and peer alcohol use strongly correlated with self-reported use of alcohol and the influence of the parental pattern did not decrease from 7th to 10th grade in a study of 688 white Houston students.

54) Barnes (1986) did random sample face-to-face interviews of 125 families given $25 for participation. Abstaining mothers had the most abstaining children and abstaining fathers had the fewest heavy drinking children. Heavier drinking mothers had adolescents with the highest rate of heavier drinking and heavier drinking fathers also had a disproportionate rate of heavier drinking teens. Maternal support was strongly linked to fewer teen. High paternal support was also associated with fewer alcohol problems but not as strongly.drinking problems. Parental control was less clearly associated with high or low drinking possible due to a mixing of angry and loving control, authoritarian and authoritative control, consistent and inconsistent control, etc.

55) Brook (1986) in a longitudinal study found use by siblings, as well as use by parents, to be predictive of initiation to alcohol use in 11th and 12th graders.

56) Burnside (1986) surveyed 2,645 Houston 7th to 10th graders and found a significant correlation between parental alcohol use and student alcohol use. Children from stepparent and single-parent families reported more alcohol use than children from intact families, but the parents in the stepparent and single-parent families also reported more alcohol use than the intact-family parents. Some increased teen use in non-intact families remained even after controlling for parental alcohol use. Parental use was classified into low user, moderate user, and heavy user. Teen alcohol use was more strongly linked to parental use than family intactness.

57) Needle (1986) studied 508 randomly selected middle-class white 11- to 13-year-olds and their peers, siblings, and parents by questionnaire and 98% were followed for three years. Substance use was very light in view of age. Peer use most closely linked with pre-teen use followed by sibling use and then parental use which were more minor after controlling for peer and sibling influence for cigarettes, alcohol, and marijuana. Friends' disapproval of cigarette, alcohol and marijuana use were more strongly correlated with lack of use followed by sibling and parental disapproval with the correlation with parental disapproval much weaker. Siblings (38%) and friends (46%) are often marijuana suppliers for teens that use. Alcohol, tobacco, and marijuana use was strongly linked to use by older siblings. This was followed by peer and parental use with peer use stronger but both significant.

58) Saltz (1986) found that in families where one or both parents drink, the children are more likely to drink.

59) Gfoerer (1987), from national NIDA 1979 and 1982 surveys, 1,117 parents-child or sibling-child pairs were selected. Gfoerer found a significant correlation between parent and child alcohol use. He also found a correlation between sibling alcohol use and the subject child's use. Strong relationships were also found between parental and sibling cigarette, alcohol, or marijuana use and teen use. Even moderate maternal and paternal alcohol use was strongly correlated with youth alcohol and marijuana use. Teen marijuana use seemed more narrowly linked to parental and sibling marijuana use rather than all three when compared to earlier studies. Even infrequent use of drugs by adults may influence teens to experiment with drugs. The author concludes, "Prevention of drug use by teenagers may be promoted by fathers, mothers, and older siblings abstaining from the use of cigarettes, alcohol, and marijuana."

60) Kandel (1987) studied 1100 teen-parent-best friend triads of New York public high school students with a questionnaire survey. Peer influence on drug and alcohol use was much stronger than parental influence. Parental alcohol use did have a significant relationship with increasing the initiation of drinking by youth and the frequency of alcohol use. Closeness to parents was important in reducing drug involvement and the frequency of alcohol use.

61) Parker (1987) in a survey of 1027 adults in metropolitan Detroit found that the offspring of heavy and very heavy drinking parents were three times more likely to be dependent problem drinkers than the offspring of abstaining, light, and moderate parents. Unfortunately, the study does not break out these latter three types of parents. 13% of the offspring of heavy drinking parents vs 4 % of the others became dependent problem drinkers. Having a blue collar occupational status was also correlated with increased dependent problem drinking.

62) A study of 341 medical students found a high occurrence of frequent and heavy drinking among 1st-year medical students. Potential problem drinkers as parental drinking increased from abstaining to light to frequent to heavy with father's drinking habits having the strongest effect. The heavy drinkers attend church infrequently. An estimated 7% of physicians become alcoholic. Forney (1988) found 13% of students abstainers, 44% light drinkers and 17% heavy drinkers. Some studies of med students have found lower rates of drinking. White and male students were more likely to be heavy drinkers.

63) In a study by Grube (1988) of 2,782 Irish 12- to 18-year-olds parental drinking was non-significantly (p<.19) less for youths who drank less after controlling for peer drinking, parental disapproval, cigarette smoking and other factors which were all significant to p<.001 (positive and negative expectancies, problem behavior) and maleness (p<.02). Since parental behaviors may influence choice of peers, but peer behaviors cannot influence who are the parents, the use of such multiple regression analysis may exaggerate in apparent influence of peer behavior and underestimate the influence of parental behavior.

64) Lubben (1988) interviewed 298 adult Filipinos and found 29% of men and 3% of women were heavy drinkers with 19% of men and 55% of women abstainers. Drinking by Filipino women was considerably less than for urban women in Los Angeles where the study was done. Higher education and being Catholic was related to increased female drinking. Increased income was related to less drinking. There was a small but non-significant trend towards a higher percentage drinking when parents drank. The small numbers in the study precluded more detailed analysis.

65) Chi (1988) interviewed 218 adult Chinese in Los Angeles of whom 89% were Asian born. 22% were abstainers and 14% heavy drinkers. This is less than in the Caucasian population. Parental drinking and going to or giving parties were the most important variables in a multivariate analysis of drinking vs. abstaining. Having parents who opposed drinking was also significant. Having friends who drink was a more significant predictor of heavy drinking as was frequenting bars and nightclubs.

Alcohol consumption in China has dramatically increased since 1946. Yen reported alcohol consumption of 9 liters per year in 1957 to 27 liters in 1984. The increase was linked to technological change and prosperity (1986)

66, 67, 68) Chi (1989) interviewed 298 Chinese, 295 Japanese and 280 Koreans in Los Angeles. Japanese-Americans were the heaviest male drinkers with 29% in the heavy category, vs. 26% of Koreans and 14% of Chinese. Male abstainers made up 17% of Japanese, 22% of Chinese and 44% of Koreans. The large majority of Japanese were American born while the Chinese and Koreans were almost all Asian born. Abstaining by women was 27% Japanese, 51% Chinese, and 75% Korean. Heavy drinking occurred in 12% of Japanese women but 0% of Chinese and 1% of Koreans. Parental drinking was a highly significant variable for Korean males and also significant for Korean females. Parental drinking was marginally significant, i.e. to the p < .10, for Japanese males and Chinese females and had only a trend toward increasing drinking in Chinese males and Japanese females. Asian males had more friends who drank, worshiped less often, and were more likely to go to bars or nightclubs. In a multivariate analysis, parental drinking was a significant factor for Koreans and having friends who drank was significant for all three groups of heavy drinkers. Heavy drinking by Japanese and Korean males was as frequent as in U.S. Caucasian populations.

69) Wilks (1989) studied 106 Australian intact families. Parental drinking and attitudes about their child's drinking and the 18- or 19-year-old offspring perception of the parental drinking and attitudes. He also studied the same-sex best friend's perceived and actual drinking as well as the friend's norms. Strongest predictors of alcohol use for males were their perceptions of their father's and mother's drinking and their father's actual drinking. Their internalized norms were best predicted by the perception of their friends drinking. Father's, but not mother's drinking also influenced the drinking of the daughters, although their friend's norms and perceived drinking also had an impact. They conclude both parent and peer drinking have a major social influence. Youth perceptions of parental and peer drinking and attitudes was highly correlated with parental and peer reports.

70) Several studies have looked at the relationship of parental drinking and rates of drinking by their children in early or late teenage years. An Australian study of 10- and 12-year-old children is typical of these studies finding attitudinal and behavioral evidence of early drinking increased by a parental drinking role-model (Family Systems Medicine, 1990).

71) Andreasson (1991) found the number of deaths in a 20-year follow-up of 49,464 Swedish draftees from alcohol psychosis, alcoholism, alcohol intoxication, liver cirrhosis, and pancreatitis and deaths overall to increase in number when comparing those with father's who drink rarely, sometimes, or often. The lowest death rate was for non-drinking draftees, followed by light, then moderate, then heavy drinkers.

72) Casswell (1991) studied 632 New Zealand children from ages 9 to 15. Children of infrequently drinking parents and children of mothers who drank lighter amounts were more likely to be abstainers. Those told of the bad effects of alcohol were more likely to be an abstainer than those given a liberal message. 13- and 15-year-old children of moderately drinking fathers consumed the most alcohol the last occasion they drank and the frequency with which children drank was very significantly increased by mother's and father's increased frequency of drinking. 90% of New Zealand adults are drinkers and this proportion is already established in children by age 15. Already at age 9, 80% of children drinking a little, most sipping but not drinking from their parents' cups. Casswell (1982) had previously reported in a less detailed study slightly higher levels of problem drinking in teens who thought their parents were strongly opposed to drinking vs those who set only some limits. In that study, teens who perceived their parents as not minding if and how they drank had clearly the highest rates of problem drinking. Parents who set only some limits vs. didn't like teen drinking or who were strongly opposed to teen drinking clearly had the highest percentage of teen drinking overall.

73) A prospective study of 3917 9th and 10th graders at 23 Australian high schools found sons of heavy drinking fathers drank more heavily. Unfortunately, the levels of drinking by abstinent, light, and moderate drinking parents and sons is not broken out (Cumes-Rayner, 1991). The heavy drinking boys were more impulsive, neurotic, left school earlier, and antisocial.

Walker (1978) suggested that parental attitudes toward adolescent drinking seem not only to affect drinking patterns, but may also be the best predictor of adolescent drinking.

74) Ellickson (1991) followed 1189 southern California 7th graders for one year. An incredible 76% were already alcohol users. Parental and peer alcohol use played a significant and important role in the initiation and frequency of drinking in the previous non-drinkers as did peer marijuana use. Peer and parental approval of alcohol use were also significant factors. Of those who were already users at the beginning of the study, parental, sibling, and peer alcohol use still played a role in increasing the offers of both alcohol and cigarettes, the expectation of using alcohol and the seeing of alcohol at not harmful. Parental use did play a role in increasing excessive or binge drinking as well. Exposure to illicit drugs, primarily marijuana, accelerated heavy drinking behavior.

75) In a study by Faulkner of 108 fraternity pledges, students from families who held a view of total abstinence consumed only an average of 54 ounces of alcohol during the month while those those from families who permitted heavy drinking consumed an average of 413 ounces. The mean of the remaining pledges was 120 ounces. Interestingly, those who took there first drink within the family setting consumed 175 ounces vs. 84 ounces for those consuming their first drink with friends of the same age. Only four of the students had never drank. Consumption was considerably lower in those who attended church very frequently (79 ounces) vs frequently (113) and only moderately (157).

76) Forney (1991) in an anonymous questionnaire study of 1,177 blacks in grades 6, 8, 10, and 12 in Georgia and South Carolina, two-thirds urban and one-third rural father's and mother's drinking behavior both very strongly correlated with the teen's drinking behavior (p<.0001). The same was also true to the best friend's drinking behavior. Although 75% had tried alcohol, 53.4% were currently abstainers. 34% had their first drink under the age of 11. Oddly, those living with both natural parents had a higher than expected number of heavy drinkers, while those living with a single parent had a lower than expected number of abstainers. Abstaining parents tended to have abstaining children. Students who drank more had more knowledge about alcohol. Mother's work status (73% working full or part-time) did not affect drinking nor did inner-city vs. rural residence. Those who abstained or drank lightly were more conservative in their attitudes even with the "acceptable" uses of alcohol such as drinking for religious purposes, during special family events, or at mealtime with the family. Although the black rates are lower than white rates in the same study, the author expresses concern about the sizeable amount of black drinking (7.2% heavy, 11.5% moderate, 28% light), especially in view of the higher cirrhosis rates in blacks. Unfortunately, although drinking behavior was categorized as abstainer, light, moderate, and heavy, these categories are not reported separately in the paper (Medical College of Georgia).

77) Green (1991) interviewed 846 Scottish 15-year-olds and their parents with a postal questionnaire follow-up one year later. The children of smoking parents were much more likely to smoke that those of non-smoking parents. The odds of drinking by young people at age 16 was increased if either mother drank or both parents. A statically positive association existed between young females drinking and parental drinking where both parents drank. There was also a positive and significant association between parents' and young people's drinking in non-manual households. The influence of the levels of parental drinking on the levels on youth drinking are not reported nor was the rate of problem drinking.

78) Dull (1992) in a study of 557 incoming California State University freshman found maternal alcohol use followed by peer use major predictors for alcohol use. Peer alcohol use and sibling alcohol abuse and to a lesser degree maternal alcohol use contributed to marijuana use in the students. Paternal use, and paternal and maternal alcohol abuse failed to contribute to the prediction equation. Paternal use, however, was strongly correlated with the student alcohol use and weakly correlated (p<.05) with marijuana use. Paternal alcohol abuse was not at all correlated to alcohol use in the offspring. Parental use was a stronger predictor of student use than peer use.

79) Iannotti (1992) studied 4th and 5th graders for drug use. The perception of friends' use, perceptions of family use, and and actual use by classmates were good predictors of substance use. Peer attitudes and behaviors were more influential for socially censured behaviors such as using alcohol without parental permission than for using alcohol with parental permission.

Tec (1974) observed that altering the use of alcohol and other drugs by adults is prerequisite to changing the substance use patterns of adolescents.

Alcohol and addictive substances are widely available, openly advertised, and regularly stocked in many homes (Akers, 1977). More juveniles are introduced to alcohol at home than anywhere else (Kirk, 1979). When Fontane and Layne (1977) asked subjects to identify the situations in which their drinking behavior is encouraged most by social pressures, home was mentioned more often than any place other than parties.

A survey of 137 teetotalers vs 465 other English church members by Hughes (1985) did report teetotalers were more likely to report blood relatives with alcohol problems (54% vs 33%) and more likely report drinking problems in children although the authors state the numbers were too small to permit a valid comparison and did not report the actual figures. Also, the problems were determined by interviews with the church members and not with the relatives or children, leaving the possibility that teetotalers are more likely to note and remember problems related to drinking by relatives than are drinkers.

Of 27 studies looking at heavy drinking in offspring, other substance abuse linked to parental alcohol use or alcohol problems in offspring, 26 find a harmful effect of parental drinking on their children with increases in alcohol abuse by the children.

80) Cohen (1993) studied 940 7th grade parent-child pairs. Child alcohol and tobacco use most strongly associated with risk-taking behaviors and friend's drug use, low grades, and parental alcohol use. Time spent with parents, communication, positive relations with parents and amount of time spent without adult supervision affected teen alcohol and tobacco use. Parents tended to underestimate the number of child's friends who use drugs, their child's risk-taking behaviors, and the amount of unsupervised time. Parental concern about their child's appearance was also a predictor. Cohen, Deborah A (1993). Can parents predict whether their seventh-grade children use drugs? USC Dept Family Medicine 1420 San Pablo, LA.

Most children are initially introduced to alcohol under the supervision of their own parents (Davies, 1972; Harford, 1983).

Thomas E. Radecki, M.D., J.D.

modern-psychiatry.com