Ethnic/Racial Differences in Adolescent Prescription Drug Misuse (PDF)

Original Article Ethnic/Racial Differences in Peer and Parent Influence on Adolescent Prescription Drug Misuse Bridgid M. Conn, MA, Amy K. Marks, PhD ABSTRACT: Purpose: To describe ethnic/racial group differences in prescription drug misuse within a na- tionally represe...

Original Article Ethnic/Racial Differences in Peer and Parent Influence on Adolescent Prescription Drug Misuse Bridgid M. Conn, MA, Amy K. Marks, PhD ABSTRACT: Purpose: To describe ethnic/racial group differences in prescription drug misuse within a na- tionally representative sample of US adolescents. Also to identify potential sociocultural influences on this health risk behavior. Methods: A secondary data analysis was conducted on the public-use data of the 2010 National Survey of Drug Use and Health using the records of 12- to 17-year-old African American, Hispanic, and white participants (N 5 18,614). Logistic and Poisson regression analyses focused on examining the predictive role of perceived parental and peer substance use disapproval on adolescents’ prescription drug misuse and how these social influences varied by ethnic group. Results: Within this sample, 10.4% of ado- lescents endorsed misusing 1 or more subtypes of prescription drugs. The results showed significant ethnic group differences in rates of prescription drug misuse such that white adolescents reported the highest rates and African American adolescents reported the lowest rates of prescription drug misuse. Rates of misuse for Hispanic adolescents fell in between. Importantly, perceived parental and peer substance use disapproval decreased the likelihood of prescription drug misuse most significantly among white adolescents compared to Hispanic and African American adolescents. Furthermore, influence of disapproval differed by the type of substance, indicating ethnic group differences in disapproval, such as views of alcohol versus marijuana use. Conclusions: These findings provide new ethnic group–specific information about the role that the attitudes of peers and parents on substance use may play in whether adolescents misuse prescription drugs. Future studies should explore possible parent/peer-related socialization mechanisms, which may account for these ethnic group differences. (J Dev Behav Pediatr 35:257–265, 2014) Index terms: adolescents, substance abuse, prescription drugs, health risk, ethnic minority health. C urrently, misuse and abuse of prescription drugs is one of the fastest growing drug epidemics in the United Rx.3” Despite slight decreases in the NMUPD among adults, rates among adolescents seem to remain stable, States, particularly among adolescents.1 Misuse is com- warranting attention regarding the reasons US adoles- monly defined as both the use of prescription medi- cents remain active in NMUPD.4 cations without a doctor’s prescription and use simply In an effort to begin to understand which group of for the “experience or feeling the drug caused.2(p.5)” In adolescents may be at the greatest risk for prescription 2006, the number of US youth older than 12 years who drug misuse, significant differences in the rates of initiated nonmedical use of prescription opioids within NMUPD between white and ethnic minority adolescents the span of a year was greater than the estimated number have been documented consistently in the recent litera- of youth who initiated either illicit cannabis or cocaine ture.5,6 Specifically, white adolescents reported higher use during that same period.2 Significant increases in the rates of NMUPD than their non-white peers.7 In their nonmedical use of prescription drugs (NMUPD) have led review, Young et al7 also noted several peer (e.g., peer to labeling today’s American adolescents as “Generation attitudes) and parental factors (e.g., parental bonding/ disapproval of drug use) associated with adolescent From the Department of Psychology, Suffolk University, Boston, MA. NMUPD; however, a few studies have attempted to ex- Received November 2013; accepted January 2014. plore how such factors may explain these racial/ethnic This study is based on secondary data analyses conducted as part of a doctoral disparities. The current study is informed by theoretical dissertation. perspectives, which move beyond the traditional cross- Presented as a poster in the 2013 Society for Child Development Conference in Seattle, Washington, April 2013. cultural “risk” models, which document between-group A. K. Marks received grants from the Jacobs Foundation. B. M. Conn received differences without examining relevant mechanisms or funding from the American Psychological Association to attend the annual APA processes that may explain these discrepancies. In fact, conference and Psychology Summer Institute. B. M. Conn is employed in the Primary several seminal articles8,9 have highlighted such issues in Children’s Hospital and received payment for lectures from the Suffolk University. clinical sciences, in particular, the use of race/ethnicity Disclosure: The authors declare no conflict of interest. as an “explanatory construct” in and of itself. Among Address for reprints: Amy K. Marks, PhD, Department of Psychology, Suffolk University, 41 Temple St, Boston, MA 02114; e-mail: [email protected]. their proposed guidelines for cultural research, Helms et al9 call for researchers to move away from using race Copyright © 2014 Lippincott Williams & Wilkins as an “explanation” and to shift focus to culture-specific Vol. 35, No. 4, May 2014 www.jdbp.org | 257 factors (e.g., values) to explain the risk behaviors in the tobacco, marijuana, and alcohol use among adolescents. development of any health issues rather than perpetuating For example, peer substance use has been related to group-level comparisons. cigarettes, marijuana, and alcohol use among US-born In a review of theoretical models on culture and hu- Hispanic and African American adolescents.16,17 Ennett man development, Garcia Coll et al8 stated that the study et al18 found that African American and Hispanic ado- of culture is vital in the confirmation, creation, and ex- lescents who reported having conversations with their pansion of theories of human development. Historically, parents about the rules and discipline related to tobacco research in the United States has taken a universal or etic use were less likely to start or increase the use of ciga- approach to studying the impact of culture on normal rettes compared to white adolescents. However, the in- and abnormal development, leading to the propagation fluence of peer and parent substance use attitudes on of “deficiency models,” in other words, using race or adolescents’ NMUPD is not currently well understood. ethnicity to explain health disparities between white and Moreover, with the differing influence of these attitudes ethnic minority youth. However, Garcia Coll et al8 make on nonprescription drugs substances, such as tobacco a strong case for examining culture-based environmental and alcohol, across adolescent racial/ethnic groups, little influences as processes rather than a backdrop for de- research has been done in the way of examining how velopmental psychopathology. Thus, studies within these social influences translate to prescription drug one culture or between cultures can provide vital misuse. new information regarding the strength of cultural fac- Using data from the 2010 National Survey on Drug tors on the increased or decreased likelihood of the Use and Health, we investigated the influence of per- development of the phenomena of interest. Taking ceived parental and peer disapproval of various sub- a within-culture, or emic, perspective may more effec- stance use on NMUPD among African American, tively document how particular cultural values, beliefs, Hispanic, and white adolescents. Our primary goal was or practices specifically impact the cause and course of to answer 3 questions: (1) Are there differences in rates developmental psychopathology.8 Thus, the current of NMUPD among ethnic/racial groups in this sample? study takes a culture-specific approach to understanding (2) Does perceived peer and parental substance use NMUPD, extending previous research. More specifically, disapproval influence adolescents’ NMUPD? (3) If so, we aim to examine one aspect of cultural socialization, what are the observed differences in peer and parental parent and peer attitudes, on adolescent NMUPD. influence patterns across these ethnic groups? Methodologically, we use a within-group (or parallel analysis)10 approach to allow for meaningful cross- METHODS cultural differences in socialization patterns to emerge. Procedures In adolescence, the attitudes and behaviors of peers The National Survey on Drug Use and Health and parents have profound impacts on youths’ own (NSDUH) is a national survey that uses multistage area attitudes and behaviors related to drug use. In the pres- probability sampling methods to select a representative ent literature, there is less information on peer disap- sample of the US civilian, noninstitutionalized population proval compared to approval/acceptability. Disapproval aged 12 years or older in all 50 states (N 5 57,873). For as a construct does not represent the opposite end of the the purposes of improving the precision of drug use spectrum from approval and is, in fact, qualitatively dif- estimates for key subgroups, adolescents aged 12 to ferent. However, some studies have used disapproval as 17 years were oversampled. The 2010 sample (N 5 a way to capture a potential protective factor for ado- 18,614) included 12- to 17-year-old participants who lescents. For example, among a diverse sample of 8th, self-identified as African American (n 5 2,486; 14.8%), 10th, and 12th graders, perceived peer substance use Hispanic (n 5 3,273; 19.4%), or white American (n 5 disapproval, particularly at the school level, decreased 11,093; 65.8%). Approximately half of the study sample the likelihood of cigarette, alcohol, and marijuana use.11 was male (50.8%). The data collection method used in In addition, a study of predominantly African American NSDUH involves in-person interviews, incorporating adolescents found that caregiver and peer disapproval procedures that are likely to increase respondents’ was negatively associated with early substance use initi- cooperation and willingness to report honestly about ation.12 Other research with white, black, and Hispanic illicit behavior. More specifically, researchers used adolescents highlights the influence of parental disap- computer-assisted interviewing to increase the likelihood proval on adolescent substance use.13,14 A recent study of valid respondent reports of drug use behaviors.19 with an ethnically diverse sample of adolescents found Computer-assisted interviewing methodology combines that the attitudes of close friends differentially predicted computer-assisted personal interviewing (CAPI) and the risk of cigarette, alcohol, and marijuana use. The audio computer-assisted self-interviewing (ACASI). authors recommended further studies to better un- However, ACASI is designed to provide the respondent derstand cultural processes that may influence adoles- with a highly private and confidential means of respond- cent substance use behaviors.15 ing to questions, and it is used for questions of a sensitive To date, ethnic group examinations of these in- nature, such as substance use.20 For more sensitive fluential social factors have been limited to studies of questions pertaining to drug use, respondents listened to 258 Ethnic Differences in Prescription Drug Misuse Journal of Developmental & Behavioral Pediatrics prerecorded questions through headphones and entered disapproval was contingent upon respondents’ indicated responses directly into the computer without interviewer presence of at least 1 parent but does not assume con- observation or assistance. In addition, all identifying in- stant cohabitation. formation was kept separate from survey responses and Perceived Peer Disapproval of Substance Use respondents switched from ACASI to CAPI mode for The following independent items were used to mea- interviewers when they completed the questions. sure perceived peer substance use disapproval: “How do you think your close friends would feel about you Measures smoking 1 or more packs of cigarettes a day?” “How do Comparisons of using ACASI within NSDUH have you think your close friends would feel about you trying shown that it reduces reporting bias.21 Therefore, vari- marijuana or hashish once or twice?” “How do you think ables used in this study are considered to be based on your close friends would feel about you using marijuana valid self-reports. A reliability study was conducted to or hashish once a month or more?” “How do you think assess the reliability of responses to the 2010 NSDUH your close friends would feel about you having 1 or 2 questionnaire, and the results provide support for con- drinks of an alcoholic beverage nearly every day?” These tinued use of the survey items.22 For racial/ethnic items were based on a 3-point scale: 1 5 neither approve grouping, adolescents were asked to select a monoracial nor disapprove; 2 5 somewhat disapprove; 3 5 strongly category (e.g., non-Hispanic white) or a multiracial cat- disapprove. Similarly, responses were recoded to a di- egory, non-Hispanic more than 1 race. Of note, 4% of chotomous variable: somewhat disapprove and strongly adolescents in this sample selected the multiracial cate- disapprove 5 1; neither approve nor disapprove 5 2. gory, identifying with more than one ethnic/racial label. Biracial African American adolescents who may have also Covariates identified as Hispanic were not included in the multira- Variables identified in previous research on ethnic cial group. In the present study, our sample is composed disparities in health behaviors and that exhibited signif- of adolescents who selected monoracial categories: non- icant racial/ethnic differences in the current sample Hispanic black/African American, non-Hispanic white, were treated as covariates.23 These included socio- and Hispanic. demographic characteristics, such as gender (e.g., 1 5 Nonmedical Use of Prescription Drugs male), age (e.g., 1 5 12–13 years old; 2 5 14–15 years Self-reported lifetime nonmedical use of stimulants, old), and total family income (e.g., 1 5 ,$10,000; 2 5 sedatives, tranquilizers, and pain relievers were coded $10,000–$19,999; 3 5 $20,000–$29,999). There was either “0” (i.e., no reported use) or “1” (i.e., yes reported a significantly higher proportion of white adolescents use). Initial analyses revealed significant associations whose reported family level income above $75,000 between nonmedical use of prescription drugs (NMUPD) (x2(4, N 5 16,852) 5 2642.56; p ,.01) and whose and demographic variables (e.g., age, ethnic/racial reported poverty status was $200% above the federal group). However, the number of responses per pre- poverty threshold (x2(6, N 5 16,852) 5 2610; p ,.01) scription drug subtype was insufficient to run full model as compared to Hispanic and African American adoles- logistic regressions for each ethnic/racial group by sub- cents. In the NSDUH interview, adolescents were asked if type independently; therefore, responses for all 4 drug they would like a parent (or other adult household mem- subtypes were summed to create a composite score and ber) to answer questions about income and insurance. recoded to a dichotomous variable (i.e., 0 5 no NMUPD; Approximately 87% of adolescents in NSDUH 2010 opted 1 5 reported lifetime misuse of at least 1 type of pre- to have an adult answer these items rather than answer scription drug) for use in the full logistic regression models. themselves24; thus, this variable is generally considered to Perceived Parent Disapproval of Substance Use be an accurate assessment of household income. The following independent items were used to mea- sure perceived parental substance use disapproval: “How Analyses do you think your parents would feel about you smoking We observed demographic distributions and preva- 1 or more packs of cigarettes per day?” “How do you lence rates of NMUPD for Hispanic, African American, think your parents would feel about you trying marijuana and white adolescents. Next, bivariate associations were or hashish once or twice?” “How do you think your examined with x2 tests for categorical variables (e.g., parents would feel about you using marijuana or hashish gender). Finally, we conducted a series of Poisson and once a month or more?” “How do you think your parents logistic regressions for each ethnic group to predict the would feel about you having 1 or 2 drinks of an alcoholic likelihood of NMUPD based on perceived parental and beverage nearly every day?” These items were based peer substance use disapproval. Each item (described on a 3-point scale: 1 5 neither approve nor disapprove; above) was entered hierarchically into the regression 2 5 somewhat disapprove; 3 5 strongly disapprove. For model with socioeconomic and gender variables entered the purposes of analysis, responses were recoded to in the first step and perceived substance use disapproval a dichotomous variable: somewhat disapprove and items entered in the second step. Parent and peer mod- strongly disapprove 5 1; neither approve nor disapprove els were analyzed separately. As we ran analyses sepa- 5 2. Reported perception of parental substance use rately by ethnic/racial group, unweighted data were Vol. 35, No. 4, May 2014 © 2014 Lippincott Williams & Wilkins 259 used. Results are presented using odds ratios as the ethnic/racial group for the majority of income levels. The measure of effect size. percentage of white adolescents increased with each income bracket and was significantly greater than Afri- RESULTS can American and Hispanic adolescents for households Descriptive Statistics earning $75,000 or greater (p ,.01). Older adolescents Among the total adolescents in the 2010 National Sur- were significantly more likely to report prescription drug vey on Drug Use and Health (NSDUH) sample, 10.4% misuse compared to younger adolescents. For example, endorsed nonmedical use of 1 or more prescription drug older white adolescents were more than twice as likely subtypes (i.e., pain relievers, stimulants, sedatives, and to report tranquilizer misuse compared to their younger tranquilizers). Across these subtypes, adolescents gener- counterparts. Finally, girls were more likely to report ally reported greater nonmedical stimulant use than non- prescription drug misuse than boys, regardless of age. medical sedative, tranquilizer, and pain killer use Table 2 provides odds ratios (ORs) of adolescents’ combined. Chi-square analyses of nonmedical use of pre- NMUPD based on perceived close friends’ and parents’ scription drugs (NMUPD) revealed significant differences substance use disapproval by substance type according across ethnic groups specifically for stimulant (x2(2, N 5 to adolescents’ gender and age. 360) 5 12.01; p ,.01) and tranquilizer prescription drug Our analyses revealed distinct ethnic differences in misuse (x2(2, N 5 495) 5 44.89; p ,.01). White ado- the nonmedical use of specific types of prescription lescents reported significantly higher rates of misuse for drugs. Consistent with previous research, white adoles- stimulants (2.4%) and tranquilizers (3.4%) compared to cents were significantly more likely to report NMUPD, either African American or Hispanic adolescents. Con- specifically of both pain relievers (OR 5 1.1; confidence versely, African American adolescents reported the lowest interval [CI], 1.01–1.27) and tranquilizers (OR 5 2.2; CI, rates of stimulant (1.3%) and tranquilizer (0.9%) misuse of 1.78–2.76). In other words, white adolescents were all 3 groups. Reported misuse of stimulants (2.1%) and more than twice as likely to report nonmedical tran- tranquilizers (2.9%) among Hispanic adolescents fell be- quilizer use compared to non-white adolescents. Con- tween the rates of misuse for white and African American versely, African American adolescents were almost 500% adolescents. Table 1 shows correlations between NMUPD less likely to report nonmedical use of tranquilizers than by subtype and demographic characteristics in the their white and Hispanic peers. Additionally, although NSDUH adolescent sample. white adolescents were 70% more likely to misuse pain relievers than non-white adolescents, African American Regression Analyses adolescents were 18% less likely to misuse pain relievers Contrary to previous findings, as income increased, than their non-African American counterparts. the total likelihood of NMUPD decreased. Adolescents who reported higher family income were significantly Perceived Parental Disapproval of Substance Use less likely to endorse misuse of pain relievers, tranquil- Table 3 shows the odds of NMUPD based on per- izers, and sedatives. Stimulant misuse was not related ceived parental disapproval by substance subtype across to adolescents’ reported family income. In terms of the ethnic/racial groups. White adolescents who en- income level, adolescents did not significantly differ by dorsed stronger perceived parental disapproval of Table 1. Spearman’s Rho Correlation Coefficients of Nonmedical Prescription Drug Use (and by Type), Poverty Level, and Income Indicators, and Demographic Characteristics (i.e., Age, Gender, Ethnicity) Variables 1 2 3 4 5 6 7 8 9 10 11 12 1 Nonmedical pain reliever use — 2 Nonmedical tranquilizer use.433* — 3 Nonmedical stimulant use.310*.311* — 4 Nonmedical sedative use.179*.189*.191* — 5 Total family income 2.048* 2.039* 2.012 2.036* — 6 Poverty level 2.038* 2.032* 2.007 2.031*.846* — 7 Age.138*.120*.086*.017**.002.014 — 8 Gender.030*.029*.024*.012 2.017** 2.016** 2.009 — 9 Any nonmedical prescription drug.924*.514*.437*.246* 2.049* 2.044*.146*.036* — misuse 10 Hispanic ethnicity.010 2.001 2.003.010 2.237* 2.266* 2.010.008.013 — 11 White ethnicity 2.002.038*.022* 2.006.386*.394*.000 2.012 2.005 2.681* — 12 Black ethnicity 2.008 2.050* 2.025* 2.004 2.251* 2.230*.011.007 2.008 2.204* 2.577* *p ,.01, 2 tailed. **p ,.05, 2 tailed. 260 Ethnic Differences in Prescription Drug Misuse Journal of Developmental & Behavioral Pediatrics Table 2. Age and Gender Differences in Logistic Regression Analyses Predicting Prescription Drug Misuse Based on Perception of Parents’ and Close Friends’ Substance Use Disapproval Among White, Hispanic, and African American Adolescents b SE Wald Odds Ratio (95% Confidence Interval) White Parental disapproval Age.66.04 226.49** 1.94 (1.78–2.12) Gender 2.28.07 17.96** 0.76 (0.67–0.86) Close friends’ disapproval Age.47.05 98.21** 1.60 (1.46–1.75) Gender 2.44.07 42.50** 0.64 (0.56–0.73) Hispanic Parental disapproval Age.56.07 56.70** 1.75 (1.51–2.02) Gender 2.13.12 1.32 0.88 (0.70–1.10) Close friends’ disapproval Age.47.08 37.14** 1.60 (1.37–1.85) Gender 2.22.12 3.65 0.80 (0.64–1.01) African American Parental disapproval Age.41.09 21.65** 1.51 (1.27–1.79) Gender 2.40.14 7.95** 0.67 (0.51–0.89) Close friends’ disapproval Age.34.09 14.32** 1.40 (1.18–1.67) Gender 2.45.14 9.84** 0.64 (0.49–0.85) **p ,.01. smoking 1 or more packs of cigarettes a day were more such strong disapproval. Additionally, white adolescents than 250% less likely to report NMUPD than their white who perceived that close friends strongly disapproved of peers who perceived less disapproval or neither ap- trying marijuana (OR 5 1.86; CI, 1.40–2.47), smoking 1 proval nor disapproval. Moreover, white adolescents or more packs of cigarettes a day (OR 5 1.33; CI, 1.08– who perceived strong parental disapproval of trying 1.63), or drinking alcohol daily (OR 5 1.25; CI, 1.02– marijuana (OR 5 2.02; CI, 1.36–2.97) and using mari- 1.53) were less likely to endorse NMUPD than those who juana monthly (OR 5 1.73; CI, 1.12–2.70) were also less did not. For Hispanic and African American adolescents, likely to report NMUPD compared to their peers. His- most of these perceived peer disapproval items did not panic adolescents who reported stronger perceived significantly predict NMUPD. However, Hispanic ado- parent disapproval of monthly marijuana use were 400% lescents who perceived disapproval (either some disap- less likely to endorse NMUPD compared to Hispanic proval or strong disapproval) of smoking marijuana adolescents who perceived lesser parental disapproval or monthly from their close friends were less likely to re- that parents neither approved nor disapproved. In addi- port NMUPD (OR 5 1.81; CI, 1.08–3.03) than those who tion, Hispanic adolescents who perceived stronger pa- reported neither approval nor disapproval. For both rental disapproval of drinking alcohol daily were 200% ethnic groups, age and gender appeared to account for less likely to endorse NMUPD than those who did not. the majority of variance compared to perceived peer Finally, African American adolescents who perceived substance use disapproval in reported NMUPD. stronger parental disapproval of drinking alcohol daily were 285% less likely to report NMUPD than those who DISCUSSION perceived lesser parental disapproval or that parents Until recently, research on nonmedical use of pre- neither approved nor disapproved. scription drugs (NMUPD) misuse among adolescents was limited to cross-cultural descriptions of prevalence. The Close Friends’ Disapproval of Substance Use present study confirmed the presence of these ethnic As presented in Table 3, white adolescents who per- group differences and novel findings that suggest that ceived close friends’ strong disapproval of monthly perceived peer and parental substance use disapproval marijuana use were more than 230% less likely to report uniquely influence adolescent NMUPD. In sum, results NMUPD compared to their peers who did not perceive were consistent with previous research7 such that white Vol. 35, No. 4, May 2014 © 2014 Lippincott Williams & Wilkins 261 Table 3. Logistic Regression Analyses Predicting Prescription Drug Misuse Based on Perception of Parents’ and Close Friends’ Substance Use Disapproval Among White, Hispanic, and African American Adolescents b SE Wald Odds Ratio (95% Confidence Interval) White Parental disapproval Smoking 11 packs of cigarettes/day.93.15 38.99* 2.52 (1.89–3.37) Trying marijuana.70.20 12.00* 2.02 (1.36–3.00) Smoking marijuana monthly.55.28 5.84* 1.73 (1.11–2.70) Drinking alcohol daily.31.18 2.91 1.36 (0.96–1.94) Close friends’ disapproval Smoking 11 packs of cigarettes/day.28.12 7.33* 1.33 (1.08–1.63) Trying marijuana.62.15 18.14* 1.86 (1.40–2.50) Smoking marijuana monthly.83.15 31.24* 2.30 (1.71–3.07) Drinking alcohol daily.22.10 4.55** 1.25 (1.02–1.53) Hispanic Parental disapproval Smoking 11 packs of cigarettes/day 2.23.34 0.45 0.80 (0.41–1.55) Trying marijuana 2.52.54 1.01 0.59 (0.21–1.67) Smoking marijuana monthly 1.40.51 7.53* 4.05 (1.49–10.99) Drinking alcohol daily.73.32 5.31** 2.01 (1.12–3.89) Close friends’ disapproval Smoking 11 packs of cigarettes/day.02.19 0.01 1.02 (0.71–1.48) Trying marijuana.43.26 2.64 1.54 (0.92–2.58) Smoking marijuana monthly.59.26 5.04** 1.81 (1.08–3.03) Drinking alcohol daily.15.18 0.67 1.17 (0.81–1.66) African American Parental disapproval Smoking 11 packs of cigarettes/day.04.40 0.10 1.04 (0.48–2.27) Trying marijuana.26.55 0.23 1.30 (0.44–3.84) Smoking marijuana monthly 2.50.58 0.74 0.61 (0.20–1.89) Drinking alcohol daily 1.05.47 4.91** 2.85 (1.13–7.19) Close friends’ disapproval Smoking 11 packs of cigarettes/day.38.23 2.69 1.47 (0.93–2.32) Trying marijuana 2.00.32 0.00 0.98 (0.54–1.86) Smoking marijuana monthly.48.32 2.19 1.61 (0.86–3.05) Drinking alcohol daily.07.25 0.07 1.07 (0.66–1.73) *p ,.01, **p ,.05. adolescents demonstrated the highest rates of NMUPD increased risk for NMUPD, it does not fully account for compared to African American and Hispanic adolescents. the observed differences among African American, His- As previous studies of adolescent alcohol, tobacco, and panic, and white adolescents. Previous studies posit that illicit drug use have similarly demonstrated,2,25,26 in the lower rates of NMUPD among Hispanic and African present study, African American adolescents demon- American adolescents may be related to their decreased strated the lowest rates of NMUPD, whereas rates of likelihood of being prescribed medications in the first NMUPD among Hispanic adolescents fell between those place.26 Other research suggests that messages from the of the other groups. We observed that higher income immediate environment (e.g., home), such as accept- was related to lower adolescent NMUPD misuse and that ability of drug use, may significantly influence adolescent African American adolescents were significantly less substance use behaviors.27 likely than both white and Hispanic adolescents to mis- The present study provides preliminary support for use prescription drugs. Thus, these results seem to sug- the notion that substance use disapproval by important gest that while income is an important factor in the socialization agents, such as family and peers, may 262 Ethnic Differences in Prescription Drug Misuse Journal of Developmental & Behavioral Pediatrics significantly influence NMUPD. Interestingly, despite higher rates of receiving prescriptions for certain types having the highest rates of NMUPD, white adolescents of medications from doctors.33 These trends may also be who reported the highest levels of parental substance due to lower rates of specific drug treatments among use disapproval across all substances demonstrated girls compared to boys (e.g., for attention-deficit hyper- significantly less risk of NMUPD compared to both activity disorder),34 leading girls to self-medicate with African American and Hispanic adolescents. However, diverted medications from family or peers.35 These find- ethnic minority parental substance use attitudes differ- ings are highlighted for future research as understanding- entially predicted adolescent NMUPD according to mediating factors that contribute to gender differences in substance type. For example, stronger perceived pa- NMUPD may provide useful in tailoring intervention and rental disapproval of daily alcohol use predicted de- prevention efforts. Finally, older age was also a consistent creased likelihood of NMUPD among African American factor increasing the likelihood of NMUPD among all adolescents, whereas among Hispanic adolescents, adolescents regardless of ethnic/racial background. strong perceived parental disapproval of monthly mari- Several limitations in this study should be considered. juana predicted decreased likelihood of NMUPD. Thus, it First, because the current study used an existing national may be that ethnic minority parents promote unique database, analyses were limited in the variables that messages about these licit and illicit substances, which could be examined and by participants’ response may then influence how their adolescents view and choices. For example, the perceived substance use dis- subsequently misuse prescription drugs.27 approval item responses were restricted to “strongly” or Conversely, perceived peer substance use disapproval “somewhat” disapprove or “neither approve nor disap- was generally not as significant a predictor of NMUPD prove.” Using close-ended responses limits our ability to among ethnic minority adolescents, although perceived assess nuances of such messages about substance use. peer disapproval of substance use significantly decreased More specifically, importantly, the 2010 National Survey the likelihood of NMUPD among white adolescents. on Drug Use and Health (NSDUH) did not assess per- While perceived peer substance use has been found to ceived parental and peer disapproval of NMUPD. In significantly influence certain behaviors (e.g., cigarette addition, the items did not assess for perceived peer or smoking and polydrug use) among urban, low-income, parental approval of substance use; thus, it is recom- ethnically diverse adolescents,28 to our knowledge, this mended that future studies endeavor to measure the full is the first study to examine differences in peer and spectrum of attitudes regarding substance use. Second, parental influences on NMUPD among ethnic/racial mi- while the use of a national data set increases generaliz- nority adolescents. Our findings add support to growing ability, the results cannot be generalized to all adolescents, evidence that parents continue to remain a vital part of particularly those in other ethnic or racial groups not adolescents’ decision making, particularly regarding included in this study. In addition, the present study did potentially risky behaviors,29 and that these influences not include adolescents who self-identified as multiracial, may be significant and unique in non-white adolescents. a segment of the population that is rapidly increasing in Some researchers posit that this lasting influence is an the United States. As the fastest growing youth group, it artifact of the core beliefs, values, and practices instilled will be vital to understand how multiple cultural identities in children by their parents.30–32 Moreover, our results and influences may shape the development of health- raise questions regarding the mechanisms or processes related behaviors for multiracial, multicultural youth in through which these ethnic group differences may arise. our nation.36,37 Third, there are well-documented limi- As white adolescents were most significantly influenced tations in the use of self-report surveys versus empirical or by parental and peer attitudes, future studies should aim observed data.38 Therefore, even though the NSDUH was to understand factors that place this group at a higher designed to eliminate such bias through the use of risk for NMUPD compared to ethnic minority adoles- a computer-assisted interview,19,20 we cannot say with cents. Conversely, researchers should examine the so- certainty that no bias was introduced during data collec- ciocultural factors that would explain lower rates of tion. Fourth, the current study was also limited by the use NMUPD among Hispanic and African American adoles- of cross-sectional data. In future research, longitudinal cents compared to white adolescents. It is just as im- data are needed to further examine the direction of these perative to begin to understand culture-specific factors observed relationships between parental and peer sub- (e.g., parental health behavior socialization) that may act stance use attitudes and adolescent NMUPD. to protect ethnic minority adolescents from engaging in Although our sample contained sufficient participants NMUPD. Such research may greatly inform theoretical in each ethnic group to conduct analyses, the number of perspectives leading to data-driven, culturally informed white adolescents far exceeded that of African American NMUPD prevention and treatment initiatives. and Hispanic adolescents combined. Moreover, the small Finally, our results provide further support of gender number of ethnic minority participants barred us from and age trends in adolescent NMUPD.7 Specifically, girls being able to examine these predictive relationships were more likely to report NMUPD than boys. Several between different classes of prescription drug (e.g., reasons are posited for this gender difference. For in- stimulant vs opioid). Thus, greater representation of stance, girls may report higher rates of NMUPD due to ethnic minorities within national samples of adolescents Vol. 35, No. 4, May 2014 © 2014 Lippincott Williams & Wilkins 263 would be beneficial for future studies. Finally, although 10. Cooper CR. Multiple selves, multiple worlds: cultural perspectives NSDUH may provide potential best estimates of NMUPD on individuality and connectedness in adolescent development. In: Master A, ed. Cultural Processes in Child Development: Minnesota in the community, in future studies, it may also be Symposium on Child Development. Mahwah, New Jersey: helpful to examine this phenomenon among acute ado- Lawrence Erlbaum; 1999:25–57. lescent populations, which may allow for closer exami- 11. Kumar R, O’Malley PM, Johnston LD, et al. Effects of school-level nation of possible developmental pathways toward or norms on student substance use. Prev Sci. 2002;3:105–124. away from NMUPD. 12. Fite PJ, Wynn P, Lochman JE, et al. The influence of neighborhood disadvantage and perceived disapproval on early substance use Despite these limitations, our study provides mean- initiation. Addict Behav. 2009;34:769–771. ingful new information about ethnic group differences in 13. Sargent JD, Dalton M. Does parental disapproval of smoking adolescent NMUPD. The results also highlight an im- prevent adolescents from becoming established smokers? portant distinction in understanding the influence of Pediatrics. 2001;108:1256–1262. parents versus peers during the formative period of ad- 14. Wood MD, Read JP, Mitchell RE, et al. Do parents still matter? Parent and peer influences on alcohol involvement among olescence. If parents hold a greater amount of influence recent high school graduates. Psychol Addict Behav. 2004;18: among adolescents, particularly regarding NMUPD, an 19–30. important next step would include qualitative inquiry to 15. Mason MJ, Mennis J, Linker J, et al. Peer attitudes effects on further explore the aspects of parental socialization that adolescent substance use: the moderating role of race and gender. shape substance use behavior. Furthermore, it will be Prev Sci. 2014;15:56–64. important to elucidate possible mediators, such as par- 16. Elkington KS, Bauermeister JA, Zimmerman MA. Do parents and peers matter? A prospective socio-ecological examination of ticular cultural beliefs or practices, which may influence substance use and sexual risk among African American youth. prescription drug use/misuse among ethnic minority J Adolesc. 2011;34:1035–1047. adolescents in protective ways. In sum, our results pro- 17. Prado G, Huang S, Schwartz SJ, et al. What accounts for differences vide support for the prevention and treatment initiatives in substance use among U.S.-born and immigrant Hispanic aimed at educating parents to a greater extent about the adolescents? results from a longitudinal prospective cohort study. J Adolesc Health. 2009;45:118–125. use of prescription drugs and how their messages about 18. Ennett ST, Bauman KE, Foshee VA, et al. Parent-child various substances may influence adolescents’ drug- communication about adolescent tobacco and alcohol use: what do related behaviors. Taking such measures may signifi- parents say and does it affect youth behavior? J Marriage Fam. cantly contribute to efforts to address this growing 2001;63:48–62. health concern among our nation’s youth. 19. Turner CF, Ku L, Rogers SM, et al. Adolescent sexual behavior, drug use, and violence: increased reporting with computer survey technology. Science. 1998;280:867–873. REFERENCES 20. Simoes AA, Bastos FI, Moreira RI, et al. A randomized trial of audio 1. Compton WM, Volkow ND. Abuse of prescription drugs and the computer and in-person interview to assess HIV risk among drug risk of addiction. Drug Alcohol Depend. 2006;83:S4–S7. and alcohol users in Rio De Janeiro, Brazil. J Subst Abuse Treat. 2. Substance Abuse and Mental Health Services Administration. A Day 2006;30:237–243. in the Life of American Adolescents: Substance Use Facts. 21. Gfroerer JC, Tan LL. Substance use among foreign-born youths in Substance Abuse and Mental Health Services Administration, the United States: does the length of residence matter? Res Prac. Office of Applied Studies. Rockville, MD; 2007. 2003;93:1892–1895. 3. Partnership for a Drug-Free America. Key findings of 18th Annual 22. Substance Abuse and Mental Health Services Administration. Partnership Attitude Tracking Study on teen drug abuse 2005: Reliability of Key Measures in the National Survey on Drug Use “Generation Rx: A Culture of Pharming Takes Root.” 2006. and Health (Office of Applied Studies, Methodology Series M-8, Available at: http://www.drugfree.org/wp-content/uploads/2011/ HHS Publication No. SMA 09–4425). Rockville, MD; 2010. 04/Key-Findings-Overview-FINAL3.pdf. Accessed May 15, 2013. 23. Buka SL. Disparities in health status and substance use: ethnicity and 4. Substance Abuse and Mental Health Services Administration. socioeconomic factors. Public Health Rep. 2002;117:S118–S125. Prescription Drug Abuse in the Workplace. Substance Abuse and 24. LeBaron PA, McHenry G, Warren LK. Data quality of adolescent Mental Health Services Administration, Office of Applied Studies. reports on person and household level income and program Rockville, MD. Available at: http://workplace.samhsa. participation. 2012. Presented as a poster at the 67th Annual gov/pdf/prescription%20drug%20abuse%20fact%20sheet.pdf. American Association for Public Opinion Research Conference, Accessed February 7, 2012. Evaluating New Frontiers in Public Opinion and Social Research, 5. Peralta RL, Steele JL. Nonmedical prescription drug use among US Orlando, FL, May 2012. college students at a Midwest university: a partial test of social 25. Johnston LD, O’Malley PM, Bachman JG, et al. Monitoring the learning theory. Subst Use Misuse. 2010;45:865–887. Future National Results on Adolescent Drug Use: Overview of 6. Teter CJ, McCabe SE, Cranford JA, et al. Prevalence and motives for Key Findings. (NIH Publication No. 07–6202). Bethesda, MD: illicit use of prescription stimulants in an undergraduate student National Institute on Drug Abuse; 2007. sample. J Am Coll Health. 2005;53:253–262. 26. Herman-Stahl MA, Krebs CP, Krouti LA, et al. Risk and protective 7. Young AM, Glover N, Havens JR. Nonmedical use of prescription factors for methamphetamine use and nonmedical use of medications among adolescents in the United States: a systematic prescription stimulants among young adults aged 18 to 25. Addict review. J Adolesc Health. 2012;51:6–17. Behav. 2006;32:1003–1015. 8. Garcia Coll C, Akerman A, Cicchetti D. Cultural influences on 27. Nasim A, Corona R, Belgrave F, et al. Cultural orientation as developmental processes and outcomes: implications for the study a protective factor against tobacco and marijuana smoking for of development and psychopathology. Dev Psychopathol. 2000;12: African American young women. J Youth Adolesc. 2007;36:503– 333–356. 516. 9. Helms JE, Jernigan M, Mascher J. The meaning of race in 28. Epstein JA, Bang H, Botvin JG. Which psychosocial factors psychology and how to change it: a methodological perspective. moderate or directly affect substance use among inner-city Am Psychol. 2005;60:27–36. adolescents? Addict Behav. 2007;32:700–713. 264 Ethnic Differences in Prescription Drug Misuse Journal of Developmental & Behavioral Pediatrics 29. Groff P, Morrish J, Smith V. Parental influence on adolescent risk 34. Angold A, Erkanli A, Egger HL, et al. Stimulant treatment for behaviours: a strategy to empower parents. Inj Prev. 2012;18(suppl children: a community perspective. J Am Acad Child Adolesc 1):A108–A109. Psychiatry. 2000;39:975–984. 30. Belgrave FZ, Chase-Vaughn G, Gray F, et al. The effectiveness of 35. McCabe SE, Boyd CJ, Young A. Medical and nonmedical use of a culture-and gender-specific intervention for increasing resiliency prescription drugs among secondary school students. J Adolesc among African American preadolescent females. J Black Psychol. Health. 2007;40:76–83. 2002;26:133–147. 36. Humes KR, Jones NA, Ramirez RR. Overview of Race and Hispanic 31. Belgrave FZ, Townsend TG, Cherry VR, et al. The influence of an Origin: 2010. 2010 Census Briefs. U.S Department of Commerce. Africentric worldview and demographic variables on drug Washington, DC: U.S. Department of Commerce, U.S. Census knowledge, attitudes and use among African American youth. Bureau; 2011. J Community Psychol. 1997;25:421–433. 37. Saulny S. Census Data Presents Rise in Multiracial Population of 32. Resnicow K, Baranowski T, Ahluwahlia J, et al. Cultural sensitivity Youths. 2011. Available at: http://www.nytimes.com/2011/03/25/ in public health: defined and demystified. Ethn Dis. 1999;9:10–21. us/25race.html?_r52&src5tptw. Accessed December 29, 2013. 33. Murray ML, de Vries CS, Wong ICK. A drug utilisation study of 38. Kandel DB, Schaffran C, Griesler PC, et al. Salivary cotinine antidepressants in children and adolescents using the General concentration versus self-reported cigarette smoking: three patterns Practice Research Database. Arch Dis Child. 2004;89:1098–1102. of inconsistency in adolescence. Nicotine Tob Res. 2006;8:525–537. Vol. 35, No. 4, May 2014 © 2014 Lippincott Williams & Wilkins 265