Valerie S. Harder

Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research

There is considerable interest in using propensity score (PS) statistical techniques to address questions of causal inference in psychological research. Many PS techniques exist, yet few guidelines are available to aid applied researchers in their understanding, use, and evaluation. In this study, the authors give an overview of available techniques for PS estimation and PS application. They also provide a way to help compare PS techniques, using the resulting measured covariate balance as the criterion for selecting between techniques. The empirical example for this study involves the potential causal relationship linking early-onset cannabis problems and subsequent negative mental health outcomes and uses data from a prospective cohort study. PS techniques are described and evaluated on the basis of their ability to balance the distributions of measured potentially confounding covariates for individuals with and without early-onset cannabis problems. This article identifies the PS techniques that yield good statistical balance of the chosen measured covariates within the context of this particular research question and cohort.

International comparisons of behavioral and emotional problems in preschool children: Parents' reports from 24 societies

International comparisons were conducted of preschool childrens behavioral and emotional problems as reported on the Child Behavior Checklist for Ages 1½–5 by parents in 24 societies (N = 19,850). Item ratings were aggregated into scores on syndromes; Diagnostic and Statistical Manual of Mental Disorders–oriented scales; a Stress Problems scale; and Internalizing, Externalizing, and Total Problems scales. Effect sizes for scale score differences among the 24 societies ranged from small to medium (3–12%). Although societies differed greatly in language, culture, and other characteristics, Total Problems scores for 18 of the 24 societies were within 7.1 points of the omnicultural mean of 33.3 (on a scale of 0–198). Gender and age differences, as well as gender and age interactions with society, were all very small (effect sizes < 1%). Across all pairs of societies, correlations between mean item ratings averaged .78, and correlations between internal consistency alphas for the scales averaged .92, indicating that the rank orders of mean item ratings and internal consistencies of scales were very similar across diverse societies.

Adolescent cannabis problems and young adult depression: Male-female stratified propensity score analyses

Cannabis use and depression are two of the most prevalent conditions worldwide. Adolescent cannabis use is linked to depression in many studies, but the effects of adolescent cannabis involvement on young adult depression remain unclear and may differ for males versus females. In this cohort study of youth from a mid-Atlantic metropolitan area of the United States, repeated assessments from 1985 (at age 6 years) through 2002 (at age 21 years) were made for 1,494 individuals (55% female). Measured covariate differences between individuals with and without cannabis problems were controlled via propensity score techniques. The estimated risk of young adult depression for adolescents with cannabis problems was not significantly different from that for comparison adolescents for either females (odds ratio = 0.7, 95% confidence interval: 0.2, 2.3) or males (odds ratio = 1.7, 95% confidence interval: 0.8, 3.6). The evidence does not support a causal association linking adolescent-onset cannabis problems with young adult depression.

COMT Val158Met Genotype as a Risk Factor for Problem Behaviors in Youth

OBJECTIVE To test the association between the catechol-O-methyltransferase (COMT) Val158Met polymorphism and both aggressive behavior and attention problems in youth. We hypothesized that youth carrying a Met allele would have greater average aggressive behavior scores, and that youth exhibiting Val-homozygosity would have greater average attention problems scores. METHOD Complete data on maternally rated Child Behavior Checklist (CBCL) measures of aggressive behavior (AGG) and attention problems (AP), COMT polymorphism data, demographics, and maternal parenting quality were available for 149 youth (6 to 18 years old). Multivariable linear regression models were used to test the degree to which youth COMT Val158Met genotype was associated with AGG and AP while statistically controlling for age, gender, parental socioeconomic status (SES), and maternal parenting quality from the Alabama Parenting Questionnaire. RESULTS Mothers of Met-carriers rated their children higher on average AGG scores when compared with mothers of Val-homozygotes (p = .016). Further analyses revealed that this association was even more robust for maternal ratings of direct aggression (p = .007). The hypothesized association between Val-homozygosity and higher average AP scores relative to average AP scores of Met-carriers did not quite reach statistical significance (p = .062). CONCLUSIONS After controlling for demographics, SES, and maternal parenting quality as confounders, there remains a strong association between youth carrying a Met allele and higher average AGG scores relative to Val-homozygotes.

Drinking and stress: An examination of sex and stressor differences using IVR-based daily data

BACKGROUND Research on the relation of stress to alcohol consumption is inconsistent regarding the direction of effects, and this association has been shown to vary by sex and type of stress. We sought to build upon the stress-drinking literature by examining the direction of the stress-drinking association over time as well as sex and stressor differences using daily data. METHOD 246 heavy drinking adults (67% men) aged 21-82 reported daily stress levels and alcohol consumption over 180 days using Interactive Voice Response (IVR). Baseline daily hassles were examined as an alternative measure of stress. Generalized estimating equations (GEEs) were conducted to test the stress-drinking association accounting for alcohol dependency at baseline and sex and stressor type as moderators. RESULTS IVR daily stress predicted increased alcohol consumption the following day, whereas baseline level of daily hassles did not. Examining the opposite direction of effects, IVR ratings of daily alcohol consumption predicted decreased next-day stress. Stress predicted higher alcohol consumption the next day for men but there was no significant association for women. For both sexes, drinking predicted decreased stress the next day, but this effect was stronger for women. CONCLUSIONS This study generally supported the drinking to cope and self-medication hypotheses, with findings that increased stress led to increased drinking. The time-varying relation between stress and alcohol appears to be sex- and measure-specific, however. Therefore, interventions targeted at stress management found to be effective for one sex should not be presumed to be applicable to the other.

Syndromes of Self-Reported Psychopathology for Ages 18–59 in 29 Societies

This study tested the multi-society generalizability of an eight-syndrome assessment model derived from factor analyses of American adults’ self-ratings of 120 behavioral, emotional, and social problems. The Adult Self-Report (ASR; Achenbach and Rescorla 2003) was completed by 17,152 18–59-year-olds in 29 societies. Confirmatory factor analyses tested the fit of self-ratings in each sample to the eight-syndrome model. The primary model fit index (Root Mean Square Error of Approximation) showed good model fit for all samples, while secondary indices showed acceptable to good fit. Only 5 (0.06%) of the 8,598 estimated parameters were outside the admissible parameter space. Confidence intervals indicated that sampling fluctuations could account for the deviant parameters. Results thus supported the tested model in societies differing widely in social, political, and economic systems, languages, ethnicities, religions, and geographical regions. Although other items, societies, and analytic methods might yield different results, the findings indicate that adults in very diverse societies were willing and able to rate themselves on the same standardized set of 120 problem items. Moreover, their self-ratings fit an eight-syndrome model previously derived from self-ratings by American adults. The support for the statistically derived syndrome model is consistent with previous findings for parent, teacher, and self-ratings of 1½–18-year-olds in many societies. The ASR and its parallel collateral-report instrument, the Adult Behavior Checklist (ABCL), may offer mental health professionals practical tools for the multi-informant assessment of clinical constructs of adult psychopathology that appear to be meaningful across diverse societies.

Incremental Validity of Test Session and Classroom Observations in a Multimethod Assessment of Attention Deficit/Hyperactivity Disorder

This study tested the incremental validity of behavioral observations, over and above parent and teacher reports, for assessing symptoms of Attention Deficit/Hyperactivity Disorder (ADHD) in children ages 6 to 12, using the Test Observation Form (TOF) and Direct Observation Form (DOF) from the Achenbach System of Empirically Based Assessment. The TOF Attention Problems and DOF Intrusive scales contributed significant unique variance, over and above parent and teacher ratings, to predicting parent and teacher ratings of hyperactivity and impulsivity and predicting categorical diagnoses of ADHD-Combined type versus Non-ADHD and ADHD-Combined type versus ADHD-Predominantly Inattentive type. The TOF Oppositional and Attention Deficit/Hyperactivity Problems scales contributed unique variance to predicting parent ratings of hyperactivity and impulsivity and the DOF Oppositional and Attention Deficit/Hyperactivity Problems scales contributed unique variance to predicting teacher ratings of hyperactivity and impulsivity.

Antipsychotic Medication Prescribing in Children Enrolled in Medicaid

BACKGROUND: Although the increase in treatment of children and adolescents with antipsychotic medications has been well documented, much less is known about the factors related to the use of these agents and how closely the treatment follows best practice recommendations. METHODS: Prescribers of each antipsychotic medication prescription issued for a Medicaid-insured child in Vermont aged <18 years were sent a prior authorization survey that assessed several domains including the clinical indication, other treatments, metabolic monitoring, prescriber specialty, and prescription origin. These variables were combined to categorize prescriptions as following indications approved by the US Food and Drug Administration (FDA) and best practice guidelines. RESULTS: The response rate of the survey was 80%, with 677 surveys from 147 prescribers available for analysis; more than one-half of the respondents were primary care clinicians. Overall, the clinical indication for an antipsychotic prescription followed best practice guidelines 91.7% of the time, with overall best practice guidelines followed at a rate of 50.1%. An FDA indication was followed in 27.2% of cases. Psychiatrists were significantly more likely to follow best practice guidelines than nonpsychiatrists. Antipsychotic medications were typically used only after other pharmacologic and nonpharmacologic treatments were ineffective, although previous treatment with cognitive-behavioral therapy was uncommon (15.5%). Metabolic monitoring that included serial laboratory tests was reported in 57.2% of cases. CONCLUSIONS: Current prescribing patterns of antipsychotic medications for children and adolescents follow best practice guidelines approximately one-half of the time, with nonadherence often related to lack of metabolic monitoring.

Multiple Traumas, Postelection Violence, and Posttraumatic Stress Among Impoverished Kenyan Youth

Research on posttraumatic stress disorder (PTSD) among youth has focused on specific subgroups from developed countries. Most of the worlds youth and war-like violence, however, is concentrated in developing countries, yet there is limited mental health data within affected countries. This study focused on a random community-based sample of 552 impoverished youth ages 6-18 within an informal settlement in Nairobi, Kenya, which experienced war-like violence for a month following the contested presidential election of 2007. Six months after the violence ended, 99 (18%) had PTSD according to the UCLA PTSD Reaction Index (Steinberg, Brymer, Decker, & Pynoos, 2004), and an additional 18 (3%) were found to have partial PTSD due to high overall scores. Kenyan psychologists conducted diagnostic interviews and found the positive predictive value of the assessment tool to be 72% in this sample; the confirmed prevalence was 12%. Similar to other studies worldwide, Criterion C (avoidance) was the limiting factor for diagnosing PTSD according to the DSM-IV-TR, and parent-child agreement was at best fair. The number of traumatic experiences was strongly associated with PTSD outcomes. Differences due to age or sex were not found. The findings indicate the need for universal mental health services for trauma-exposed youth and their families in the impoverished informal settlements of Nairobi, Kenya.

Cocaine Use and Educational Achievement: Understanding a Changing Association Over the Past 2 Decades

Trends in cocaine use over the past 2 decades were compared across levels of education in a population-based US sample of adults. Significant inverse associations between educational achievement and cocaine use after 1990 were driven by dramatic decreases in persistent cocaine use among more highly educated adults, whereas persistent cocaine use remained relatively unchanged among those who did not finish high school. This emerging health disparity highlights the need for improved interventions that target persistent cocaine users with low educational achievement.