Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

The International Journal of the Addictions, 27(10), 1197-1209, 1992

Family Structure and Adolescent Risk-Taking Behavior: A Comparison of Mexican, Cuban, and Puerto Rican Americans Jan S. Sokol-Katz, M.A.,' and Patricia M. Ulbrich, Ph.D.*Y*

' Department of Sociology, University of Miami, Coral Gables, Florida 33124 Department of Sociology, The University of Akron, Akron, Ohio 44325

ABSTRACT We use the Hispanic HANES to examine whether family structure is related to alcohol and drug use among Mexican, Puerto Rican, and Cuban American adolescents. Mexicans adolescents living in female headed households have higher rates of drinking, drug use, and overall risk-taking behaviors than those living with both parents. Puerto Ricans adolescents living in female-headed households have higher rates of overall risk-taking behaviors than those living with both parents. Family structure is unrelated to Cuban adolescent risktaking behavior. There is no evidence that gender modifies the effect of family structure for adolescent risk-taking behaviors.

Key words. Drugs; Alcohol; Hispanic; Adolescent; Family structure

*To whom all correspondence should be addressed. I197

Copyright 01992 by Marcel Dekker, Inc.

SOKOL-KATZ AND ULBRICH

I198

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

INTRODUCTION The increase in adolescent alcohol and drug use during the 1980s is undisputed (Kaplan et al., 1982; Matsueda, 1982; Mills and Noyes, 1984; Johnson et al., 1987; Tonkin, 1987). In their study of high school students, Johnson et al. (1987) report that nearly all seniors had tried alcohol (92%) and that more than half of seniors had tried some illicit drug (61%). There is a positive correlation between marijuana and other illicit drugs (Yamaguchi and Kandel, 1984) and between marijuana and alcohol use (Johnson et al., 1987). The increase in alcohol and drug use coupled with the tendency to use multiple substances has rekindled interest in the study of adolescent risk-taking behaviors. One explanation for adolescent alcohol and drug use focuses on family structure. We have a “two-tiered system of childrearing” in the United States (Tienda and Angel, 1982). One tier consists of two-parent households in which domestic and economic roles are shared by two adults. The second tier consists of single-parent households in which the entire burden of economic and domestic life is often borne by one adult. According to social control theory, the family provides an external social structure that contains the behavior of family members (Reckless, 1962; Umberson, 1987). Family members are likely to engage in risk-taking behaviors when this containment system is absent. In this view, single-parent families may be less effective than two-parent families in controlling adolescent risk-taking behaviors. Studies of adolescent alcohol and drug use provide some support for these propositions (Jensen, 1972; Kandel et al., 1978). Much of this research is based on non-Hispanic White populations. Studies of Hispanic adolescent substance use typically categorize their study group as “Hispanic” regardless of the Hispanic subgroup under study (Humm-Delgado and Delgado, 1983). However, Hispanics are not a homogenous group. There may be important structural and cultural factors influencing variations in substance use within this broad category (Adlaf et al, 1989; McBride and Page, 1980). For example, Puerto Ricans have a higher proportion of female-headed households and lower average household incomes than Cubans or Mexicans (Perez, 1985, 1986). If we are to understand the nature and extent of substance use among Hispanic adolescents, we must examine multiple indicators of adolescent risk-taking behaviors across Hispanic subgroups. This article is guided by three research questions aimed at assessing the relationship between family structure and the extent of risk-taking behaviors among Mexican, Cuban, and Puerto Rican adolescents. First, do adolescents living in a two-parent household report less alcohol or drug use than those living in single-parent households? Family relationships provide an external social structure that contains the behavior of family members. The basic prem-

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

FAMILY STRUCTURE AND ADOLESCENT BEHAVIOR

1199

ise of social control theory is that human beings engage in risk-taking behavior because norm violation is attractive and exciting (Traub and Little, 1985). If family roles discourage risk-taking behaviors (Reckless, 1962; Umberson, 1987), the lack of family roles should increase the probability of risk-taking behaviors. It is in this sense that the two-tiered system of child rearing in the United States has potentially serious implications for the emotional and social well-being of children (Hirschi, 1983; Angel and Worobey, 1988a, 1988b). The single parent, usually a woman, has responsibility for maintaining a household. The lack of a partner deprives women of financial and emotional support at the same time it gives them sole responsibility for childrearing. In other words, single-parent households have limited resources to control or contain adolescent risk-taking behaviors. There is some evidence that Hispanic adolescents in single-parent households engage in more risk-taking behavior than their counterparts in two-parent households. Research involving Mexican adolescents suggests that divorce and a single-parent family structure are related to delinquency (Martinez et al., 1979) and inhalant abuse (Cohen, 1973; Bonnheim and Korman, 1985). Similarly, Szapocznik et al. (1989) found that Cuban adolescent males in treatment for drug use are often living in families disrupted by divorce or abandonment. One interpretation of these findings is that the lack of social control associated with single-parent families is a contributing factor for risk-taking behaviors among Hispanic adolescents. The second question concerns gender differences in risk-taking behaviors. Whatever the family structure, there is some evidence that parents selectively impose controls on sons and daughters (Hagan, 1989). Daughters are controlled more extensively than sons. Hagan (1989) argues that girls will be less delinquent than boys because extensive parental control reduces daughters’ preferences for risk-taking behaviors. Indeed, research suggests that males not only consume more alcohol than females (Forney et al., 1988; Johnson et al., 1987), but males are also more likely than females to use drugs (Fidell, 1982). Finally, are boys living in female-headed households more likely than girls or boys living in two-parent households to use alcohol or drugs? If daughters are controlled mostly by mothers (Hagan, 1989), girls may be less affected than boys by the loss of parental control. Consequently, the relationship between family structure and alcohol or drug use may be conditioned by gender. In other words, alcohol or drug use should be higher among boys than girls in female-headed households. We use the Hispanic Health and Nutrition Examination Survey (HANES) to examine both specific and overall risk-taking behaviors of adolescents in three Hispanic groups. First, we describe the alcohol and drug use of Mexican, Cuban, and Puerto Rican adolescents. Then we examine whether family structure is related to the extent of risk-taking behaviors among Mexican, Cuban, and Puerto Rican adolescents.

I200

SOKOL-KATZ AND ULBRICH

DATA AND METHODS

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

Data The data for this study come from the Hispanic Health and Nutrition Examination Survey (National Center for Health Statistics, 1985). We use three subsamples consisting of Mexican, Puerto Rican, and Cuban-Americans aged 12 to 18 who were part of their respective sampling universe, identified themselves as part of that Hispanic subgroup, and completed both the interview and the physician examination. In addition, the unique structure of the Hispanic HANES enabled us to match adolescents with their mothers. If mothers were married and living with their spouse, they were defined as a two-parent family. If mothers were not married and there were no other adults living in the household, they were defined as a single-parent family. Adolescents living in other family structures were excluded. Based on these criteria, there are 144 Cuban, 299 Puerto Rican, and 794 Mexican-American adolescents.

Measures Respondents were asked if they had drunk any alcohol during the past month. Alcohol is then measured in two ways. Alcohol1 contrasts ( I ) those who drank alcohol (beer, wine, or liquor) and (0) those who did not drink. Alcohol2 is the mean daily amount of alcohol (ethanol) consumed in the reference period. Using Tukey’s 1977 ladder of reexpression, we transformed the mean daily amount of alcohol by taking the inverse of its negative square root [ l/d-(alcohol)] in order to reduce its skewness. Respondents were also asked whether they use marijuana, barbiturates, inhalants, or cocaine. If a respondent indicated that s/he had taken a particular drug, it was scored as a “1.” Drugs is an additive index of marijuana, barbiturate, inhalant, and cocaine use. Factor analysis revealed that each drug loaded on one factor for Mexicans and Cubans, and that all drugs except barbiturates loaded on one factor for Puerto Ricans. In order to make comparisons across groups, we use the total score on drugs for all groups. The range is from 0 to 4. Risk-taking behavior is an additive index of alcohol1 and drugs. The range is 0 to 5 . Family Structure contrasts (0) single-parent and (1) two-parent households. Sex is scored (0) male and (1) female. Age is respondent’s age in years. Family income is combined family income from all sources during the past 12 months. It is scored as an interval measure ranging from $500 to $50,500 or above. Location is measured by dichotomous variables differentiating whether the respondent lived in the city, suburb, or rural area. In all groups, City contrasts whether the respondent lived in a ( I ) city or (0) not. For Mexicans only,

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

FAMILY STRUCTURE AND A9OLESCENT BEHAVIOR

1201

Suburb contrasts whether they live in a (1) suburb or (0) not; and Rural contrasts whether they live in a (1) rural area or (0) not. Language is an indicator of acculturation. Respondends had a choice of participating in either English or Spanish. Language of interview is scored (0) English and (1) Spanish. Mother’s employment status is a dichotomous variable which contrasts (1) employed and (0) not employed. Mother’s physical health status is scored (1) excellent, (2) very good, (3) good, (4) fair, and (5) poor.

Analysis We present means and percentages of all variables in the three Hispanic groups. Then we present regression analyses with alcohol consumption, drug use, and risk-taking behavior as the dependent variables. In addition to examining the additive effects of the independent variables, we test for the interaction effect of sex and family structure. This allows us to determine whether the relationship between sex and risk-taking behaviors is significantly different among male and female adolescents in these household structures. Because the Hispanics HANES is based on a complex sample design, it was necessary to incorporate sample weights for proper analyses of the data (National Center for Health Statistics, 1985). We use SURREGR to compute regression coefficients and their corresponding standard errors by taking into account the complex sample design (Kovar and Johnson, 1986).

FINDINGS Descriptive Table 1 shows the distribution of alcohol and drug use as well as adolescent and family characteristics in each group. In these data, over 25 % of Mexican and Puerto Rican adolescents report having used drugs. Moreover, the proportion who report drug use is twice as great as the proportion who report alcohol use. In contrast, roughly equal proportions of Cuban adolescents report alcohol and drug use (16.1 and 15.0%). The vast majority of Puerto Rican adolescents (73.2%) but a slight minority of Mexican and Cuban adolescents live in female-headed households (49.3 and 42.6%). Boys and girls are equally represented in each group. They are 15 years old, on average. There are marked differences in adolescent and family characteristics between these groups. For example, a greater proportion of Cubans than Mexicans or Puerto Ricans took the interview in Spanish. Not only do Puerto Ricans have a higher proportion of female-headed households than Mexicans or Cubans, but a small proportion of Puerto Rican mothers are employed and their household incomes are low. These sociocultural patterns provide different contexts for adolescents risk-taking behaviors.

SOKOL-KATZ A N D ULBRlCH

1202

Table 1. Breakdown of Alcohol Use, Drug Use, Individual, and Family Characteristics by Hispanic Group

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

Mexicans Alcohol Drugs Single parent household Males Family income Age Spanish Language Urban Mothers employed Mother’s health N

Puerto Ricans

Cubans

11.70%

16.10%

49.30%

28.40% 73.20%

42.60%

47.00%

50.80%

49.00%

$17,476 14.71 years 44.90%

$12,535 15.07 years

$18,658 14.99 years 82.50%

51.60% 42.90%

85.60% 23.30%

29.00% 54.90%

3.32

3.58

3.03

154

284

136

14.70% 29.70%

63.40%

15.00%

Regression Analysis The multiple regression of alcohol consumption in Table 2 shows that family structure has a significant effect for alcohol use among Mexicans, but not among Puerto Ricans or Cubans. Mexican adolescents living in single-parent households report more alcohol use than those living in two-parent households. As expected, males report significantly more alcohol consumption than females in each group. Age has a significant positive relationship to alcohol use in each group, with older adolescents reporting more alcohol use than younger adolescents. Language, place of residence, and family income are related to alcohol use in some but not all groups. Cuban adolescents who took the interview in English reported more alcohol use than those who took the interview in Spanish. Puerto Rican adolescents living in a city report more alcohol consumption than those living in suburbs. Finally, Mexican adolescents living in households with higher incomes report more alcohol use than those in lower income households.

I203

FAMILY STRUCTURE AND ADOLESCENT BEHAVIOR

Table 2. Standardized Regression Coefficients of Alcohol Consumption on Individual and Family Characteristics for Mexican, Puerto Rican, and Cuban Adolescents (standard error in parentheses)

Subst Use Misuse Downloaded from informahealthcare.com by Chulalongkorn University on 01/06/15 For personal use only.

Mexicans Family structure Sex Age Family income Language City Suburb R2 Intercept

N

*

Family structure and adolescent risk-taking behavior: a comparison of Mexican, Cuban, and Puerto Rican Americans.

We use the Hispanic HANES to examine whether family structure is related to alcohol and drug use among Mexican, Puerto Rican, and Cuban American adole...
661KB Sizes 0 Downloads 0 Views