!!!!!!

Title&

Below&please&list&the&title&of&this&resource.&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

! The!Impact!of!Social!Support!and!Self:Esteem!on!Adolescent!Substance!Abuse!Treatment! Outcome!! ! ! Author&

Below&please&list&the&author(s)&of&this&resource."

! Stephanie!S.!Richter,!Sandra!A.!Brown!and!Mariam!A.!Mott! ! Citation&
Below&please&cite&this&resource&in&APA&style.&For&guidance&on&citation&format,&please&visit& http://owl.english.purdue.edu/owl/resource/560/01/&

! Richter,!S.!S.,!Brown,!S.!A.,!&!Mott,!M.!A.!(1991).!The!impact!of!social!support!and!self:! ! esteem!on!adolescent!substance!abuse!treatment!outcome.!Journal"of"Substance" " Abuse,"3(4),"371:385.!! ! ! Summary&

Below&please&provide&a&brief&summary&of&this&resource.&If&an&abstract&is&available,&feel&free&to&copy&and&paste&it&here.&

! ! Although!poor!social!resources!and!low!self:esteem!have!been!Implicated!in!the! development!of!teenage!drug!abuse,!the!role!of!these!factors!in!the!remission!of!adolescent! addiction!remains!unclear!This!study!examines!social!support!characteristics!and!self: esteem!in!relation!to!outcome!following!adolescent!chemical!dependency!treatment.! Adolescents!and!their!parents!completed!self:report!questionnaires!and!a!research! interview!during!treatment!and!at!6!and!12!months!post:!treatment.!Two!types!of!outcome! were!assessed!at!follow:up:!(I)!alcohol!and!drug!use,!and!(2)!functioning!in!major!life! domains!(e!g!,!family,!school/work,!peers)!Results!indicate!that!the!quality!of!social! resources!(Ie,!drug:use!patterns!of!supports)!reported!during!treatment!was!related!to! alcohol:!and!drug:use!status!post:treatment,!with!abstainers!reporting!more!nonusing! supports!than!teens!who!returned!to!heavy!drug!use.!Self:esteem!and!the!degree!of! satisfaction!with!Social!support!during!treatment!were!negatively!correlated!with!the! number!of!major!life!problems!during!the!6!months!following!discharge.!Altogether,! inpatient!measures!of!self:esteem,!number!of!high:quality!supports,!and!social!support! satisfaction!accounted!for!16%!of!the!variance!in!6:month!substance!use!outcome!and!25%! of!the!variance!in!psychosocial!functioning!6!months!post:treatment.!Six:month!social!

!!!!!!!!!!!!!!!!!!!!!!!!!!P.!O.!Box!6448!!Reno,!NV!89513!!Phone:!760:815:3515!!staciemathewson@me.com! !

!!!!!!

support!and!self:esteem!measures!were!similarly!related!to!l:year!outcome.!The! implications!of!these!findings!for!adolescent!drug!abuse!treatment!are!discussed!!
"

!!!!!!!!!!!!!!!!!!!!!!!!!!P.!O.!Box!6448!!Reno,!NV!89513!!Phone:!760:815:3515!!staciemathewson@me.com! !

Journal of Substance Abuse, 3, 371-385 (1991)

The Impact of Social Support and Self-Esteem on Adolescent Substance Abuse Treatment Outcome
Stephanie S. Richter
Department of Veterans Affairs Medical Center, San Diego, CA

Sandra A. Brown
Department of Veterans Affairs Medical Center, San Diego, CA and Department of Psychiatry, University of California, San Diego

Mariam A. Molt
Department of Psychiatry, University of California, San Diego

Although poor social resources and low self-esteem have been Implicated in the development of teenage drug abuse, the role of these factors in the remission of adolescent addiction remains unclear This study examines social support characteristics and self-esteem in relation to outcome following adolescent chemical dependency treatment. Adolescents and their parents completed self-report questionnaires and a research interview during treatment and at 6 and 12 months posttreatment 1\\0 types of outcome were assessed at follow-up: (I) alcohol and drug use, and (2) functioning in major life domains (e g, family, school/work, peers) Results indicate that the quality of social resources (I e, drug-use patterns of supports) reported during treatment was related to alcohol- and drug-use status post-treatment, with abstainers reporting more nonusing supports than teens who returned to heavy drug use. Self-esteem and the degree of satisfaction with SOCIal support during treatment were negauvely correlated with the number of major life problems during the 6 months following discharge. Altogether, inpatient measures of self-esteem, number of high-quality supports, and social support satisfaction accounted for 16% of the variance in 6-month substance use outcome and 25% of the variance in psychosocial functioning 6 months post-treatment. Six-month SOCIal support and self-esteem measures were similarly related to l-year outcome. The implications of these fmdings for adolescent drug abuse treatment are discussed

Recent concern about the prevalence of alcohol and drug abuse among teenagers has focused considerable attention on identifying environmental, psychological, and sociological factors associated with adolescent drug-taking behavior. Mounting evidence underscores the importance of the social environThis project was supported by grants from the National Institute on Alcohol Abuse and Alcoholism (AA07033) and the Research Sen ICe of the Department of Veterans Affairs awarded to Sandra A Brown Portions of this research were presented at the Western Psychological Association Annual Convention, Los Angeles, April 1990. Correspondence and requests for reprints should be sent to Sandra A. Brown, Psychology Service (116 B), Department of Veterans Affairs Medical Center, 3350 La Jolla VIllage Drive, San DIego, CA 92161.
371

372

5.5. Richter, 5.A. Brown, and M.A. Mott

ment in the development of teenage alcohol and drug use. Research suggests that social pressure and substance-using role models are important determinants of alcohol and drug involvement among youth. Donovan andJessor (1978) report that in a national study of adolescents the problem drinkers perceived greater approval, pressure, and models for drinking than did the non problem drinkers. Social support for drug use has likewise been found to be the most powerful predictor of marijuana use among high school and college samples (jessor, Jessor, & Finney, 1973). Although the process of social involvement is not yet fully understood, a strong relationship has consistently been found between adolescent alcohol- and drug-use patterns and those of their parents and friends (e.g., Alexander & Campbell, 1967; Kandel, 1973; Smart & Fejer, 1972). In addition to modeling substance use and providing approval for drug-taking behavior, the members of an adolescent's social network may influence alcohol and drug consumption through indirect channels. Because the drinking and drug use patterns of significant others are likely to impact on qualitative features of support such as consistency and availability (see Holden, Brown, & Mott, 1988), the likelihood that an adolescent will turn to alcohol or drugs to cope with life stresses may be greater among adolescents whose social resources abuse substances. Rhodes and Jason (1990) propose that social networks, social competencies, and community resources play a critical role in offsetting the stressors of youth that might otherwise lead to substance involvement and other problem behaviors. Inconsistencies in the literature (e.g., Shedler & Block, 1990) notwithstanding, numerous studies linking poor social resources to adolescent substance abuse provide support for this model. For example, teens who use drugs have been found to perceive significantly less love from their parents than do nonusers (Streit, .Halsted, & Pascale, 1974). A negative emotional climate in the home (Adler & Lotecka, 1973) and troubled relationships with parents (e.g., Greenwald & Luetgert, 1971; Prendergast & Schaefer, 1974; Tudor, Petersen, & Elifson, 1980) also have been associated with substance use among youths. The apparent connection between unsatisfactory personal relationships and teenage drug use is consistent with the social stress model of adolescent substance abuse. Additional support for this model is provided by studies examining the role of social competencies in adolescent drug involvement. Scherer, Ettinger, and Mudrick (1972) report that a high need for social approval may place a teen at increased risk for substance use by heightening vulnerability to social pressure. Although conflicting findings have been reported (e.g., Eskilson, Wiley, Muehlbauer, & Dodder, 1986), many studies suggest that negative self-attitudes are similarly related to adolescent drinking and drug use. Teens in drug abuse treatment have been found to have significantly lower self-esteem than the general high school population (Ahlgren & Norem-Hebeisen, 1979; Svobodny, 1982). Lower levels of self-esteem also have been associated with higher levels of substance use involvement among junior high and high school students (Pandina & Schuele, 1983). Self-attitudes may thus play an important role in determining an adolescent's ability to effectively resist social pressure to drink and use drugs as well as pressure to engage in other deviant activities. While poor social resources and negative self-attitudes have been related to

Social Support, Self-esteem, and Teenage Recovery

373

adolescent drinking and drug use, the role of these variables in the process of teenage recovery from addiction remains little understood. The present study investigates the impact of social support and self-esteem on substance use outcome and psychosocial functioning following adolescent chemical dependency treatment. Those teens with the greatest drug use among their primary social supports and with the lowest social support satisfaction and self-esteem were predicted to be most likely to return to pretreatment patterns of substance abuse. The following relationships were hypothesized: (1)The number of social supports of high, moderate, and low quality (categorized according to the support's alcohol and drug use) reported during inpatient treatment will be related to teen alcohol and drug use during the 6-month follow-up period, with abstainers reporting more high-quality supports and fewer low-quality supports, compared with the reports of relapsers. (2) The degree of satisfaction with social support during inpatient treatment will be related to substance use outcome and psychosocial functioning during the 6-month follow-up period, with greater satisfaction being associated with abstinence and fewer life problems. (3) Inpatient measures of self-esteem will be related to substance use outcome and psychosocial functioning during the 6-month follow-up period, with higher self-esteem being associated with better treatment outcome and fewer problems in major life domains. (4) Quality of social support, social support satisfaction, and self-esteem measured during inpatient treatment will contribute significantly to prediction of substance use outcome and psychosocial functioning 6 months post-treatment. In our longitudinal replication, it was anticipated that 6-month measures of social support and self-esteem similarly would be related to substance use outcome and psychosocial functioning 1 year following treatment.

METHOD
Subjects One hundred sixty adolescent inpatients in substance abuse treatment (64 females, 96 males) and their parents participated. Teens were between the ages of 12 and 18 (M = 15.9, SD = 1.3) and were predominantly Caucasian (78%). They came from families whose socioeconomic status ranged from upper-lowerto upper-upper-class (M = 30.2, SD = 12.6) as measured by the Two-factor Index of Social Position (Hollingshead, 1965). Although the majority (74%) of those studied were not currently practicing their religion, 33% of the adolescents identified themselves as Protestant, 28% as Catholic, 9% reported another religious background, and 30% indicated no religious orientation. Participants were recruited from two inpatient chemical dependency treatment programs in San Diego, California. Of the adolescents initially recruited for participation in the study, 92% were followed up in the second phase of data collection, 6 months following treatment (N = 147), and 90% of eligible teens' were again assessed 1 year after leaving the treatment program (N = 137).
I All teens for whom I year had elapsed smce leaving the treatment program were considered eligible for follow-up at the I }ear time pomt.

374

5.5. Richter, 5.A. Brown, and M.A. Mott

Procedure Adolescents consecutively admitted to the treatment programs were invited to participate in the study if they lived within 50 miles of the research facility, had a parental figure (typically the teen's biological parent) willing to independently participate in the research project, and met DSM-III-R (American Psychiatric Association, 1987) criteria for substance abuse without evidencing an independent Axis I psychiatric disorder that predated the onset of alcohol or drug abuse. Ninety-eight percent of eligible teens and parents enrolled in the study. Participation was voluntary and was conducted in accordance with the ethical guidelines of the university and treatment programs involved. As part of an ongoing longitudinal study, teens completed self-report questionnaires and a confidential structured research interview (Brown, Vik, & Creamer, 1989) 2 to 3 weeks following admission to chemical dependency treatment and again at 6 and 12 months post-treatment. A separate and confidential interview was conducted with the teen's parent at each data collection time point using the resource-person version of the structured interview (Brown et al., 1989). Following data collection, parent and teen responses were compared to establish a single reliable composite of the data. In cases of discrepancy in reports, a worst-case scenario was scored so as to avoid overestimating the adolescent's functioning. Teens were paid $25 upon completion of the 6-month follow-up interview and testing, and $50 for their participation 1 year post-treatment. Parents received $10 and $20, respectively.

Initial Assessment Social and demographic data were gathered during inpatient treatment by a trained research assistant using a 60-min. structured clinical interview (Brown et aI., 1989). Interview procedures standardized by Schuckit (1988) were employed in obtaining detailed information regarding substance use among family members as well as the teen's exposure to alcohol- or drug-abusing relatives. The Customary Drinking and Drug Use Record (CDDR; Brown, Creamer, & Stetson, 1987) was utilized in obtaining adolescent alcohol- and drug-use histories during the inpatient interview. The CDDR is based on DSM-III-R criteria (American Psychiatric Association, 1987) for substance abuse and dependence, the Quantity-Frequency-Variability Drinking Index (Cahalan & Cisin, 1968), and the Index of Drug Indulgence (Lu, 1974). This instrument provides a detailed assessment of drinking and drug use patterns, dependency symptoms, and major life problems associated with alcohol and drug use. The abbreviated version of the Social Support Questionnaire (Pierce, Sarason, & Sarason, 1986) was used to assess the perceived number of social supports and the degree of satisfaction with available supports. Measures were taken during inpatient treatment and at the 6-month follow-up interview. The Social Support Questionnaire (SSQ) has been shown to be a valid and stable instrument with internal consistency estimates ranging from .96 to .98 (Sarason, Levine, Basham, & Sarason, 1983; Pierce et al., 1986). The short form has

Social Support, Self-esteem, and Teenage RKovery

375

comparable psychometric properties (Pierce et al., 1986), including similar internal reliability (range = .89 to .93) and similar correlations with measures of personality and social competence. The abbreviated SSQ consists of six items describing situations in which social support might be important to a person. For each item, respondents are instructed to list up to nine individuals to whom they can turn for support in the manner described and to indicate on a six-point Likert scale ranging from "very satisfied" (6) to "very dissatisfied" (1) the degree of contentment with support in that area. The number of different social supports is summed, and an average satisfaction rating is computed. Drinking and drug use characteristics of identified supports were available for a subsample of adolescents (N = 69 for inpatient measures and N = 106 at 6 months post-treatment). Using this additional information, social supports were divided into three categories with regard to their alcohol- and drug-use lifestyle. Individuals who neither drank nor used drugs were classified as high-quality supports. Those who drank or used drugs on an infrequent or social basis without evidencing substance-use-related problems were considered moderatequality supports. Probable or definite substance abusers were classified as lowquality supports. Self-esteem was measured during inpatient treatment and at the 6-month follow-up interview using the self-report questionnaire employed by Labouvie and McGee (1986). The questionnaire consists of 18 interrogative items about the self (e.g., "How often do you feel that you are a good person?") and has moderate internal consistency (coefficient alpha = .74). Respondents indicate their feelings about each statement on a five-point Likert scale ranging from "never" to "almost always or always." Ratings are summed to produce a single self-esteem score, with higher scores indicating greater self-satisfaction. Follow-up Assessment Adolescents and parents were independently interviewed at 6 and 12 months post-treatment. Detailed information regarding use of alcohol or drugs during the follow-up time period was obtained using a 60-min. structured clinical interview (Brown et al., 1989) and verbal administration of the CDDR at each follow-up time point. Two types of outcome were examined at 6 and 12 months post-treatment. Adolescent substance use outcome was categorized in a trichotomous fashion based on both teen and parent reports. Abstaining teens were compared with those whose post-treatment alcohol or drug use was sporadic and time limited, as well as with those who returned to more severe levels of substance use. Teens who refrained from use of any alcohol or drugs were classified as abstainers; those who used alcohol or drugs in a time-limited fashion (:5 30 days using) were classified as minor relapsers; and teens who returned to heavy substance use 30 days using) were classified as major relapsers. Of the 147 teens assessed 6 months post-treatment, 30% were abstainers, 27% were minor relapsers (M = 5.20, SD = 6.16 days using), and 43% were major relapsers (M =

376

5.5. Richter, 5.A. Brown, and M.A. Mott

110.93, SD = 50.44 days using). Of the 137 teens assessed I }'ear following treatment, 36% were abstainers, 26% were minor relapsers (M = 5.94, SD = 8.03 days using), and 38% were major relapsers (At = 145.04, SD = 54.01 days using). The second type of treatment outcome assessed was psychosocial functioning. The number of problems occurring in each of five major life domains (family, school/work, peers, emotional health, and the law) was determined by administration of an inventory that includes the areas of Jessor's ProblemBehavior Theory (e.g., Donovan & Jessor, 1978), and Newcomb and Bentler's Domain Theory (1988). Additionally, a series of questions from the follow-up interview was used to determine whether the teen had experienced interpersonal problems, periods of depression, episodes of anxiety, or other emotional problems requiring counseling or hospitalization. The inventory consists of 39 items: 7 items assess functioning in the home environment (e.g., "Did you run away from home overnight or longer?"); 9 items pertain to school- and workrelated behavior (e.g., "Have you been suspended or expelled from school?"); 8 items examine interpersonal functioning (e.g., "Have you been involved in any fights?"); 4 items evaluate emotional health (e.g., "Have you had any emotional! psychological problems requiring professional help/therapy in the past 6 months?"); and 11 items address legal problems (e.g., "Have you been involved in any illegal occupations, like fencing, selling drugs, prostitution?"). The total number of behavioral, interpersonal, and emotional problems reported is summed to create a composite measure of psychosocial functioning posttreatment. Adolescents residing in an institution (e.g., an extended residential treatment facility) for one half or more of the follow-up time period were excluded from analyses (N = 18 at 6 months post-treatment and N = 16 at the l-year follow-up) because restrictions imposed by their living situations may bias post-treatment measures of substance use and psychosocial functioning.

RESULTS

Preliminary Analyses
Analyses of variance (ANOVAs) were conducted to determine whether pretreatment sociodemographic, background, and drug use characteristics could account for observed differences across the three substance use outcome groups (abstainers, minor relapsers, and major relapsers). The outcome groups were not significantly different on any of the demographic characteristics assessed in the study (e.g., age, gender, ethnic background, socioeconomic status, religious background, religious practice). In addition, no significant differences were found across the groups on pretreatment drinking and drug use characteristics (e.g., age of first hard drug use, number of drugs tried, number of times of alcohol or hard drug use during the 3 months prior to treatment), family history of alcoholism and drug abuse, or percentage of lifetime exposure to alcohol or drug abuse within the family. ANOVAs also were conducted to determine whether the teens unavailable at

Social Support, Self-esteem, and Teenage Recovery

377

follow-up differed from adolescents for whom follow-up data were obtained. No significant differences were found on the above variables between those assessed at 6 and 12 months post-treatment and teens unavailable at the followup time points. Quality of Social Support Teens reported an average of 9.56 (SD = 3.37) social resources on the SSQ during treatment. Multivariate analyses of variance (MANOVAs) using WiIk's lambda criterion were conducted to determine whether adolescent drug use outcome groups differed in the quality of their social resource networks. The first MANOVA examined the number of high-, moderate-, and low-quality social supports reported during treatment and substance use outcome at the 6 month follow-up. The overall MANOVA was significant (F(6, 130) = 2.30, P < .05). Post-hoc univariate F-tests revealed that the overall significance was due to group differences in the number of high-quality supports (F(2, 66) = 3.60 ,p < .05). Post-hoc comparisons using Tukey's method showed that abstainers reported significantly (p < .05) more high-quality social supports (M = 4.39, SD = 3.38) than major relapsers (M = 2.50, SD = 1.44). The mean number of moderate- and low-quality supports reported during inpatient treatment by teens in each of the 6 month substance use outcome groups was as follows: Abstainers reported 4.44 (SD = 2.15) moderate-quality supports and 2.61 (SD = 2.79) low-quality supports; minor relapsers reported 3.16 (SD = 1.21) moderate-quality supports and 2.32 (SD = 1.29) low-quality supports; and major relapsers reported 4.03 (SD = 1.96) moderate-quality supports and 2.16 (SD = 1.46) low-quality supports. Six-month measures of high-, moderate-, and low-quality supports were similarly related to outcome status 1 year following treatment (F(6, 204) = 4.23, P < .01). Post-hoc univariate F-tests revealed that the overall significance was due to differences across the groups in the number of high-quality supports (F(2, 103) = 9.49, P < .01), with abstainers reporting significantly (p < .05) more high-quality supports (M = 6.16, SD = 2.56) than both minor (M = 4.44, SD = 2.76) and major (M = 3.58, SD = 2.66) relapsers. The relationship between high-quality social support and substance use outcome is illustrated in Figure I. The mean number of moderate- and low-quality supports reported at 6 months post-treatment by teens in each of the l-year substance use outcome groups was as follows: Abstainers reported 3.21 (SD = 1.73) moderate-quality supports and 1.40 (SD = 0.72) low-quality supports; minor relapsers reported 3.15 (SD = 1.46) moderate-quality supports and 1.56 (SD = 0.89) low-quality supports; and major relapsers reported 4.07 (SD = 2.20) moderate-quality supports and 1.78 (SD = 1.13) low-quality supports. Social Support Satisfaction ANOVAs were conducted to examine the relationship between social support satisfaction and substance use outcome. When ANOVAs were significant, post-

378
7

5.5. Rkhter, 5.A. Brown, and M.A. Mott

1:

'" 0

0. -::-

:::J u

0. ..:

--4 .::: .:::J _

,

r:;' C
<)

...
0

:I: .S 3

,-

eGO;

0.

Abstalners 'nor el sets )or Relapsers

..o"l;l

.....c: 2 eu :s
:::J

S '-'

7-,-- - - - - - - - - - - - - - - - - - ----,
6

Abstainers :> Minor Relapsers Abstainers :> Major Re apsers p < .05

(;

0. ......

U 5 Cl) S
:::J

.= ii
0;
r;roo
.:. 0.

]lin
.... 0

Abstainers MinorRelapsersi
jor Relapse rs :

-.:: -c o
o
In

E
2

..0 ......

:l

0 -1--

Sub

ce-u

Outcome

(one ye ar pos t-tre atment)

Figure 1. Number of high-quality social supports by substance use outcome.

Social Support, Self-esteem, and Teenage Recovery

379

Table 1. Correlation of Social Support Satisfaction and Self-esteem with Psychosocial Functioning of Adolescents following Substance Abuse Treatment
Inpatient
Social support sansfaction Self-esteem

Number of Problems 6 Months Post-treatment

-.19
- 27

p
.03
>.01

Number of Problems 6 month Follow-up

Social support satisfaction
Self-esteem
1 Year Post-treatment

p
.01 >.01

- 24
- 32

hoc comparisons using Tukey's method were conducted to determine which outcome group(s) differed significantly. No significant differences were found across the three outcome groups at the 6-month follow-up on inpatient measures of social support satisfaction (F(2, 118) = 2.17, P = .12). Significant differences were, however, found across the l-year drug use outcome groups on 6-month measures of social support satisfaction (F (2, 106) = 4.35, P< .05), with abstainers and minor relapsers reporting significantly (p < .05) greater satisfaction with their social supports (M = 5.42, SD = 0.50, and M = 5.46, SD = 0.53, respectively) than major relapsers (M = 5.05, SD = 0.85). Pearson correlation coefficients were calculated to assess the relationship between social support satisfaction and psychosocial functioning during the first and second 6-month periods following discharge from treatment. As expected, the degree of satisfaction with social support during treatment was significantly negatively correlated with the total number of problems occurring across major life domains during the initial 6-month follow-up period (r = - .19, N = 121, P < .05). This same relationship was demonstrated at the l-year follow-up; the degree of satisfaction with social support reported 6 months posttreatment was significantly negatively correlated with the total number of problems occurring between the 6-month and l-year time points (r = - .24, N = 109, P < .05). Social support satisfaction was thus associated with better psychosocial functioning in major life domains during the follow-up time periods (see Table 1). Self-esteem ANOVAs were conducted to examine the relationship between self-esteem and substance use outcome following treatment. Although no significant differences were found across the 6-month drug use outcome groups on inpatient measures of self-esteem (F(2, 108) = 1.51, P = .23), significant differences were found across the outcome groups at the I-year follow-up on fi-month measures of self-esteem (F(2, 106) = 8.35, P< .01). Post-hoc Tukey tests

380

5.5. Richter, 5.A. Brown, and M.A. Molt

indicate that abstainers and minor relapsers obtained significantly (p < .05) higher self-esteem scores (1\1 = 69.71, SD = 7.22, and M = 70.35, SD = 8.44, respectively) than major relapsers (AI = 63.62, SD = 8.38). Pearson correlation coefficients were calculated to examine the extent to which measures of self-esteem were related to subsequent psychosocial functioning. As hypothesized, the level of self-esteem during treatment was significantly negatively correlated with the total number of problems occurring across major life domains during the first 6 months post-treatment (r = - .27, N = HI, P < .01). Similarly, self-esteem scores at the 6-month follow-up were significantly negatively correlated with the number of problems occurring between 6 months and I year post-treatment (r = - .32, N = 109, P < .01). Thus, teens with greater self-esteem reported fewer problems in major life domains at each of the followup time points (see Table I). Multiple Regression Analyses Standard multiple regression analyses were performed to assess the extent to which social support measures and self-esteem predict substance use outcome. The first analysis found a significant relationship between 6-month substance use outcome group and the three inpatient measures of self-esteem, number of high quality social supports, and social support satisfaction 51) = 3.17, P < .05). Sixteen percent of the variance in 6-month substance use outcome was accounted for by the three inpatient measures. Self-esteem (sr = .09) contributed significantly to prediction of substance use outcome group (p < .05), and the three independent variables in combination contributed another .06 in shared variability (see Table 2). Similarly, I-year substance use outcome group was predicted from 6-month measures of self-esteem, number of high-quality supports, and social support satisfaction (F(3, 101) = 10.71, P < .01), with 24% of the variance in l-year substance use outcome accounted for by the 6-month measures. Both selfesteem (sr = .05) and number of high-quality supports (sr = .11) contributed significantly to substance use outcome group (p < .05 and p < .01, respectively), and the three independent variables in combination contributed another .08 in shared variability (see Table 3). Additional multiple regression analyses were conducted to examine whether post-treatment psychosocial functioning could be predicted from social support measures and self-esteem. The first analysis found a significant relationship between 6-month psychosocial functioning and inpatient measures of selfesteem, number of high-quality social supports, and social support satisfaction (F(3, 51) = 5.64, P < .01). Twenty-five percent of the variance in 6-month psychosocial functioning was accounted for by the three inpatient measures. Self-esteem (sr = .18) contributed significantly to prediction of psychosocial functioning (p < .01), and the three independent variables in combination contributed another .07 in shared variability (see Table 4). Likewise the l-year analysis indicated that psychosocial functioning could be predicted from 6-month measures of self-esteem, number of high-quality supports, and social support satisfaction (F(3, 101) = 6.16, P < .01). Fifteen

Social Support, Self-esteem, and Teenage Recovery

381

Table 2. Multiple Regression of Inpatient Measures of Social Support and Selfesteem on fi-month Substance Use Outcome
Variables High-quality social supports Social support satisfaction Self-esteem
*P< 05
B

13
-0.22 008 -0.32

sr2 (Unique)

-0.078 0072 -0.034*

.09

R2 = 16 Adjusted R2 = .11 R = .40*

Table 3. Multiple Regression of 6-month Measures of Social Support and Self-esteem on I-year Substance Use Outcome
Variables High-quality social supports
Social support

B
-0.101** -0.173 -0.024*

13
-0.33 -0.14 -0.24

sr 2 (U ruque)

.11
R2 = 24 Adjusted R2 R = .49**

satisfaction Self-esteem *p<.05 **p<01

= .22

.05

Table 4. Multiple Regression of Inpatient Measures of Social Support and Selfesteem on 6-month Psychosocial Functioning
Variables High-quality social supports Social support satisfaction Self-esteem
**p<.Ol.

B
0298 -0.938 -0.337**

13
-012 -0.15 -0.44

sr 2 (Unique)

.18

R2 = .25 Adjusted R2 = 20 R = .50**

Table 5. Multiple Regression of 6-month Measures of Social Support and Self-esteem on J-year Psychosocial Functioning
Variables High-quality social supports Social support satisfaction Self-esteem
**p<.Ol.
B

13
-0.15 -0.14 -0.27

sr 2 (Unique)

-0.225 -0907 -0.140**

.07

R2 = .15 Adjusted R2 = .13 R = .39**

382

5.5. Richter, 5.A. Brown, and M.A. Moll

percent of the variance in 6-year psychosocial functioning was accounted for by the three 6-month measures. Self-esteem (sr = .07) contributed significantly to prediction of psychosocial functioning (p < .01), and the three independent variables in combination contributed another .09 in shared variability (see Table 5).

Gender Differences
ANOVAs were conducted to examine whether gender differences exist in the relationships of interest. A series of two-way ANOVAs were conducted with 6-month outcome group and gender as independent variables for inpatient measures of the number of high-quality supports, moderate-quality supports, low-quality supports, social support satisfaction, and self-esteem. The only analysis that revealed a significant main effect for gender was the ANOVA examining the number of low-quality supports (F(l, 63) = 4.55, P < .05). Examination of group means revealed that females reported more lowquality social supports (M = 2.93, SD = 2.60) than males (M = 1.93, SD = 0.97) during inpatient treatment. Consistent with previous analyses, no significant main effect for substance use outcome was obtained (F(2, 63) = 0.21, P = 0.81), and the two-way interaction was not significant (F(2, 63) = 0.07, P = 0.93). The above described series of analyses were then repeated with I-year outcome group and gender as independent variables for 6-month measures of social support and self-esteem. No significant gender effects were obtained in any of the analyses. DISCUSSION The results of this study indicate that social resources and self-esteem are related to teenage recovery from addiction. The quality of social support reported by adolescent inpatients was related to substance use outcome following treatment, with abstainers having more supports who do not use drugs than major relapsers and minor relapsers reporting an intermediate number of nonusing supports. Further, greater satisfaction with one's social resources and higher self-esteem were associated with fewer problems in major life domains during the year following treatment. The obtained pattern of findings is, by and large, consistent with the hypotheses of the study. The previous finding of a relationship between teenagers' drinking and drug use patterns and those of their parents and friends has been extended to include recovering adolescent substance abusers following drug abuse treatment. Particularly noteworthy is the finding that the number of abstinent social supports reported during inpatient treatment is predictive of outcome over a 6-month time period. While the observed relationship between high-quality support at the 6-month follow-up and l-year drug use outcome may be at least in part a function of abstainers choosing to be around other abstainers (e.g., attending Alcoholics Anonymous meetings), significant differences across the 6-month outcome groups on inpatient mea-

Social Support,

and

Recovery

383

sures suggest the importance of high-quality support as an antecedent of successful treatment outcome. Although inpatient measures of social support satisfaction and self-esteem were not significantly different across the 6-month outcome groups, 6-month measures of social support satisfaction and self-esteem were related to l-year substance use outcome in the manner hypothesized. It is possible that crises culminating in adolescent admission to alcohol and drug abuse treatment may have uniformly reduced social support satisfaction and self-esteem, thereby restricting these measures during treatment. Inpatient measures of self-esteem, number of high-quality supports, and social support satisfaction in combination were, however, shown to have modest power as predictors of substance use outcome and psychosocial functioning at the 6-month follow-up, and the same pattern of relationships was observed between 6-month measures and I-year outcome. Social resources and self-esteem are considered crucial to establishing and maintaining the major lifestyle changes necessary for adolescent abstinence. Because the majority of adolescent relapse occurs in social settings with pressure to drink or use drugs (Brown et al., 1989), social resources for abstention may be critical to the coping process (Myers & Brown, 1990). Abstinent social resources not only model abstention but also may demonstrate viable coping strategies for adolescents in high-risk settings, thereby enhancing the perceived acceptability of these options. Such supports may further decrease relapse risk by encouraging the adolescent to avoid social pressure situations, by providing emotional support for decisions to abstain, and by reinforcing drug-refusal efforts. Self-esteem appears to playa similarly important role in the evolution of drug abuse and the maintenance of lifestyle changes following treatment. Within the framework of stress and coping models (e.g., Lazarus & Folkman, 1984), selfesteem may influence both primary and secondary appraisal. Teens with lower self-esteem may perceive high risk for relapse settings as more threatening and themselves as being less capable of generating or implementing adequate coping strategies. Furthermore, the observed relationship between self-esteem and problems in major life domains suggests that the influence of self-esteem on appraisal may be generic rather than specific to drug abuse. The finding that short-term clinical course following adolescent drug abuse treatment is not related to sociodemographic, background, and pretreatment drug use characteristics is consistent with previous reports based on similar samples of adolescents in drug abuse treatment (e.g., Brown et al., 1989). While homogeneity of the sample and restriction of the range may contribute to this result, short-term adolescent drug abuse treatment outcome appears to be most closely linked to post-treatment social and environmental characteristics (Brown, in press). The results of this study should be considered in light of certain limitations. For example, social support characteristics and self-esteem measured at one point in time were examined in relation to psychosocial functioning and substance use outcome over a subsequent 6-month time period. Because these personal characteristics may fluctuate over time, unassessed changes in social

384

5.5. Richter, 5.A. Brown, and M.A. Mott

support or self-esteem occurring during each follow-up time period may have influenced outcome. Another consideration in substance abuse research is the validity of alcoholand drug-use outcome data. A number of measures were taken in the present study to enhance the accuracy of the self-report information obtained. Participants were reassured of confidentiality, and repeat interviews were conducted by the same interviewer whenever possible. Different interviewers were used with each parent and teen pair to protect against accidental sharing of information. Following data collection, teen reports were combined with information gathered in the separate and confidential parent interview to establish a single, maximally reliable composite of the data, and alternative data sources (e.g., medical records) were consulted in cases of question. These results have implications for adolescent chemical dependency treatment programs. For example, social support characteristics (e.g., number of high-quality supports, degree of satisfaction with available support) and selfesteem measures could be utilized to identify adolescents at increased risk for relapse and/or for experiencing serious problems post-treatment. These empirical findings argue for the value of intervention efforts that focus on assisting teens (both as inpatients and following treatment) in establishing satisfying social support networks comprised by individuals leading alcohol- and drugfree lifestyles while endeavoring to enhance the adolescent's self-esteem. Strategies of teenage social network restructuring (Catalano & Hawkins, 1985), utilization of high-quality social supports in the therapeutic process, and exploration of the ways in which abstinent supports can be useful in potential relapse situations are all likely to enhance treatment success. Experiences that result in self-esteem enhancement may also prove efficacious with adolescents. Many treatment programs currently include opportunities for volunteer efforts with the disabled or elderly, physically challenging tasks, or the mastery of new skills. The relationship between the resultant self-esteem enhancement of these experiences and adolescent substance abuse treatment outcome now needs to be explored.
REFERENCES
Adler, P.T, & Lotecka, L. (1973) Drug use among high school students' Patternsand correlates. International Journal of the AddictIOns, 8, 537-548. Ahlgren, A , & Norem-Hebeisen, A.A. (1979). Self-esteem patterns distinctive of groups of drug abusing and other dysfunctional adolescents International Journal of the AddIctIOns, 14, 759-777. Alexander, C.N. ]r, & Campbell, E.Q. (1967) Peer Influences on adolescent drinking. Quarterly Journal of Studies on Alcohol, 28, 444-453. American Psychiatric Association. (1987) Dtagnostic and stausucal manual of mental disorders (3d ed., rev) Washington, DC: Author Brown, S.A. (in press). Adolescent substance abuse Recovery patterns. In] S. Baer, G A.Mariatt, & R.] McMahon (Eds ), Addictivebehmnors acrossthe life span: Prevention, treatment, and polzcyISsues Beverly Hills, CA: Sage Publications. Brown, S.A , Creamer, V A., & Stetson, B.A (1987) Adolescent alcohol expectancies in relation to personal and parental drinking patterns Journal ofAbnormal Psychology, 96, 117-121. Brown, SA, VII.., PW, & Creamer, V.A. (1989). Characteristics of relapse following adolescent substance abuse treatment. AddlCtwe Behatnors, 14, 291-300

Social Support, Self-esteem, and Teenage Recovery

385

Cahalan, D, & Cism, I.H. (1968) American drinking pracuces: Summary of Imdings from a national probabihty sample. I. Extent of drinking by population subgroups. Quarterly Journal of Studies on Alcohol, 29, 130-151 Catalano, R.F., & Hawkins, J D. (1985) Project skills: Prehminary results from theoreucally based aftercare expenment. In R.A. Sherry (Ed ), Progress m the development of cost-effecuoe treatment for drug abusers (NIDA Research Monograph 58) Rockville, MD. Department of Health and Human Services Donovan, J E., & Jessor, R. (1978) Adolescent problem drinking: Psychosocial correlates in a national sample study Journal of Studies on Alcohol, 39, 1506-1524. Eskilson, A , WIley, M.G, Muehlbauer, G , & Dodder, L. (1986) Parental pressure, self-esteem and adolescent reported deviance. Bending the twig too far. Adolescence, 21, SOl-SIS. Greenwald, B.S., & Luetgert, MJ (1971) A companson of drug users and non-users on an urban commuter college campus. lntemauonal journal of the Addictions,6,63-78. Holden, M G , Brown, S.A., & Mott, M A (1988) Social support network of adolescents: Relation to family alcohol abuse. Amencan Journal of Drug and AlcoholAbuse, 14, 487-498. Hollingshead, A.B. (1965) Two-factor Index of Soczal POSItIOn. New Haven, CT Yale University Press Jessor, R.,Jessor, S L , & Finney,]. (1973).A SOCIal psj chology of rnanjuana use: Longuudinal studies of high school and college )outh.Journal of Personality and SOCial Psychology, 26, I-IS Kandel, D B (1973) Adolescent mar!Juana usc: Role of parents and peers Science, 181, 1067-1070. Labouvie, E W, & McGee, C.R. (1986). Relation of personahty to alcohol and drug use In adolescence. Journal ofConsulting and Clinical Ps)chology, 54, 289-293 Lazarus, R.S , & Folkman, S. (1984) Stress, appraisal, and coping. New York: Springer-Verlag Lu, K H (1974) The indexmg and analysis of drug Indulgence. lntemauonal journal of theAddictIOns, 9,785-804 Myers, M.G, & Brown, SA. (1990) Coping and appraisal In potenual relapse situations among adolescent substance abusers following treatment Journal of Adolescent Chemical Dependency, I, 95-1I5. Newcomb, MD., & Bentler, P.M. (1988) Consequences of adolescent drug use. Newbury Park, CA: Sage. Pandina, RJ , & Schuele, J A. (1983) Psychosocial correlates of alcohol and drug use of adolescent students and adolescents in treatment. Journal of Studies on Alcohol, 44, 950-973. PIerce, G R, Sarason, I.G, & Sarason, B R (1986, May) Social support questionnaire short-form Psychometrics, personality correlates and theory. Paper. presented at the Western Psj chological Association convention, Seattle, WA. Prendergast, T.J ,Jr., & Schaefer, E.S. (1974) Correlates of drmking and drunkenness among highschool students. QuarterlyJournal of Studies on Alcohol, 35, 232-242. Rhodes, J E , & Jason, L.A. (1990). A SOCIal stress model of substance abuse. Journal of Consulting and Clmical Psychology, 58, 395-401. Sarason, I G , Levine, H M., Basham, R B , & Sarason, B.R. (1983) Assessmg SOCIal support: The SOCIal Support Questionnaire. Journal of Personality and SOCial Psychology, 44, 127-139 Scherer, S E ,Ettinger, R F., & Mudrick, N.J. (1972) Need for social approval and drug use.Jounzal of Consultmg and Clinical Psychology, 38, 1I8-121. Schuckit,M A (1988). Reacuons to alcohol in sons ofalcohohcs and controls Alcohohsm: Cluneal and Expenmental Research, 12, 465-470. Shedler, J., & Block,]. (1990). Adolescent drug use and psychological health. Amencan Psychologist, 45,612-630 Smart, R G, & Fejer, D (1972) Drug use among adolescents and their parents ClOSIng the generation gap in mood modificalion.Journal of AbnormalPsychology, 79, 153-160 Streit, F., Halsted, D L., & Pascale, P.J (1974) DIfferences among youthful users and nonusers of drugs based on their perceptions of parental behavior. International joumai of theAddictiOns, 9, 749-755 Svobodny, LA (1982) Biographical, self-concept and educational factors among chemically dependent adolescents. Adolescence, 17.847-853 Tudor, C.G , Petersen, D M., & Ehfson, K W. (1980) An exarnmanon of the relauonshrp between peer and parental Influences and adolescent drug use. Adolescence, 15, 783-798.