Contamination bias in the estimation of child maltreatment causal effects on adolescent internalizing and externalizing behavior problems

BACKGROUND

When unaddressed, contamination in child maltreatment research, in which some proportion of children recruited for a nonmaltreated comparison group are exposed to maltreatment, downwardly biases the significance and magnitude of effect size estimates. This study extends previous contamination research by investigating how a dual-measurement strategy of detecting and controlling contamination impacts causal effect size estimates of child behavior problems.

METHODS

This study included 634 children from the LONGSCAN study with 63 cases of confirmed child maltreatment after age 8 and 571 cases without confirmed child maltreatment. Confirmed child maltreatment and internalizing and externalizing behaviors were recorded every 2 years between ages 4 and 16. Contamination in the nonmaltreated comparison group was identified and controlled by either a prospective self-report assessment at ages 12, 14, and 16 or by a one-time retrospective self-report assessment at age 18. Synthetic control methods were used to establish causal effects and quantify the impact of contamination when it was not controlled, when it was controlled for by prospective self-reports, and when it was controlled for by retrospective self-reports.

RESULTS

Rates of contamination ranged from 62% to 67%. Without controlling for contamination, causal effect size estimates for internalizing behaviors were not statistically significant. Causal effects only became statistically significant after controlling contamination identified from either prospective or retrospective reports and effect sizes increased by between 17% and 54%. Controlling contamination had a smaller impact on effect size increases for externalizing behaviors but did produce a statistically significant overall effect, relative to the model ignoring contamination, when prospective methods were used.

CONCLUSIONS

The presence of contamination in a nonmaltreated comparison group can underestimate the magnitude and statistical significance of causal effect size estimates, especially when investigating internalizing behavior problems. Addressing contamination can facilitate the replication of results across studies.

Effects of Receiving Pain-Related Support on Psychological Well-Being: The Moderating Roles of Emotional Responses to Support

OBJECTIVES

We investigated whether receiving greater pain-related instrumental support is associated with poorer psychological well-being among chronic pain patients who report less positive (e.g., grateful) or more negative (e.g., angry) emotional responses to support.

METHODS

We conducted regression analyses, utilizing data from two waves of interviews with 152 knee osteoarthritis patients. Three indicators of psychological well-being were examined: depressive symptoms, positive affect, and negative affect.

RESULTS

Receiving greater support was associated with poorer psychological well-being at baseline, as well as higher depressive symptoms and negative affect at the 18-month follow-up, only among patients with low positive emotional responses to support. Furthermore, receiving greater support was related to poorer psychological well-being at baseline only among patients with high negative emotional responses to support.

DISCUSSION

Care recipients' less positive emotional responses to support may be a risk factor for poorer psychological well-being in both the short- and long-term, when receiving greater support.

Contamination in Observational Research on Child Maltreatment: A Conceptual and Empirical Review With Implications for Future Research

Contamination is a methodological phenomenon occurring in child maltreatment research when individuals in an established comparison condition have, in reality, been exposed to maltreatment during childhood. The current paper: (1) provides a conceptual and methodological introduction to contamination in child maltreatment research, (2) reviews the empirical literature demonstrating that the presence of contamination biases causal estimates in both prospective and retrospective cohort studies of child maltreatment effects, (3) outlines a dual measurement strategy for how child maltreatment researchers can address contamination, and (4) describes modern statistical methods for generating causal estimates in child maltreatment research after contamination is controlled. Our goal is to introduce the issue of contamination to researchers examining the effects of child maltreatment in an effort to improve the precision and replication of causal estimates that ultimately inform scientific and clinical decision-making as well as public policy.

Epigenetic age acceleration as a biomarker for impaired cognitive abilities in adulthood following early life adversity and psychiatric disorders

Background

Early life adversity and psychiatric disorders are associated with earlier declines in neurocognitive abilities during adulthood. These declines may be preceded by changes in biological aging, specifically epigenetic age acceleration, providing an opportunity to uncover genome-wide biomarkers that identify individuals most likely to benefit from early screening and prevention.

Methods

Five unique epigenetic age acceleration clocks derived from peripheral blood were examined in relation to latent variables of general and speeded cognitive abilities across two independent cohorts: 1) the Female Growth and Development Study (FGDS; n = 86), a 30-year prospective cohort study of substantiated child sexual abuse and non-abused controls, and 2) the Biological Classification of Mental Disorders study (BeCOME; n = 313), an adult community cohort established based on psychiatric disorders.

Results

A faster pace of biological aging (DunedinPoAm) was associated with lower general cognitive abilities in both cohorts and slower speeded abilities in the BeCOME cohort. Acceleration in the Horvath clock was significantly associated with slower speeded abilities in the BeCOME cohort but not the FGDS. Acceleration in the Hannum clock and the GrimAge clock were not significantly associated with either cognitive ability. Accelerated PhenoAge was associated with slower speeded abilities in the FGDS but not the BeCOME cohort.

Conclusions

The present results suggest that epigenetic age acceleration has the potential to serve as a biomarker for neurocognitive decline in adults with a history of early life adversity or psychiatric disorders. Estimates of epigenetic aging may identify adults at risk of cognitive decline that could benefit from early neurocognitive screening.

Examining the Bidirectional Nature of Loneliness and Anxiety Among Older Adults in Daily Life

OBJECTIVES

Loneliness in later life increases the risk for adverse health outcomes; however, less is known about how loneliness is maintained. Anxiety may play an important role in maintaining loneliness, but little is known about how this connection plays out over time in daily life. This study thus focused on the within-person associations between momentary loneliness and anxiety among older adults.

METHODS

Participants were 317 diverse older adults (40% Black; 13% Hispanic, mean age = 77.45 years, 67% women) systematically recruited from the Bronx, NY, who completed ecological momentary assessments 5 times daily for 14 consecutive days. Multilevel models tested bidirectional contemporaneous, momentary cross-lagged (t - 1), day-level cross-lagged (average day to end of day), and day-to-day cross-lagged associations between loneliness and anxiety. Separate sensitivity analyses controlled for concurrent overall mood valence or depressed state. Gender and mild cognitive impairment (MCI) status were tested as moderators at all timescales.

RESULTS

Significant bidirectional associations between loneliness and anxiety were found at the contemporaneous and momentary cross-lagged (t - 1) timescales. Higher average daily loneliness predicted higher end-of-day anxiety, but not vice versa. Loneliness and anxiety were not significantly associated from day to day. Sensitivity analyses revealed some associations varied depending on inclusion of either concurrent mood valence or depressed state. Neither gender nor MCI status moderated associations at any timescale.

DISCUSSION

Findings shed light on the complex temporal ordering of loneliness and anxiety in daily life and extend contemporary theoretical notions of loneliness, including the possibility of interventions that target key moments in daily life.

The mediating role of self-regulation on the link between child maltreatment and later substance use among Latinx youth

BACKGROUND

Children who experience maltreatment are at heightened risk for substance use initiation and mental health disorders later in life. Few studies have assessed the relationship between child maltreatment and substance use among Latinx youth.

OBJECTIVE

The current study assessed the potential mediating effect of three aspects of self-regulation (emotional, behavioral, and cognitive) on the association between child maltreatment and substance use and examined whether effects varied depending on maltreatment type and severity.

PARTICIPANTS AND SETTING

This study involved a random sample of 504 Latinx youth (52 % girls, 48 % boys) between the ages of 10-12 at the start of the study.

METHODS

Study hypotheses were tested through structural equation modeling and bootstrapped random errors using the R programming language.

RESULTS

Our results indicated that higher levels of child maltreatment predicted higher levels of later substance use, as mediated by emotional and behavioral dysregulation (β = 0.09, p < 0.01), but not cognitive regulation. When separating maltreatment by subtype, we found the mediating effect was present for abuse (β = 0.09, p < 0.01), but not neglect.

CONCLUSIONS

Findings contribute to our understanding of potential causal mechanisms for the association between child maltreatment and substance use for Latinx youth.

Within-person associations of optimistic and pessimistic expectations with momentary stress, affect, and ambulatory blood pressure

BACKGROUND AND OBJECTIVES

Although dispositional optimism and pessimism have been prospectively associated with health outcomes, little is known about how these associations manifest in everyday life. This study examined how short-term optimistic and pessimistic expectations were associated with psychological and physiological stress processes.

METHODS

A diverse sample of adults (N = 300) completed a 2-day/1-night ecological momentary assessment and ambulatory blood pressure (ABP) protocol at ∼45-minute intervals.

RESULTS

Moments that were more optimistic than typical for a person were followed by moments with lower likelihood of reporting a stressor, higher positive affect (PA), lower negative affect (NA), and less subjective stress (SS). Moments that were more pessimistic than typical were not associated with any affective stress outcome at the following moment. Neither optimism nor pessimism were associated with ABP, and did not moderate associations between reporting a stressor and outcomes.

DISCUSSION

These findings suggest that intraindividual fluctuations in optimistic and pessimistic expectations are associated with stressor appraisals.

Parent-Focused Sexual Abuse Prevention: Results From a Cluster Randomized Trial

This study tested whether a child sexual abuse (CSA) prevention program, Smart Parents-Safe and Healthy Kids (SPSHK), could be implemented as an additional module in evidence-based parent training and whether the added module might detract from the efficacy of the original program. In a cluster randomized trial, six community-based organizations were randomized to deliver Parents as Teachers (PAT) with SPSHK (PAT+SPSHK) or PAT as usual (PAT-AU). CSA-related awareness and protective behaviors, as well as general parenting behaviors taught by PAT were assessed at baseline, post-PAT, post-SPSHK, and 1-month follow-up. Multilevel analyses revealed significant group by time interactions for both awareness and behaviors (ps < .0001), indicating the PAT+SPSHK group had significantly greater awareness of CSA and used protective behaviors more often (which were maintained at follow-up) compared to the PAT-AU group. No differences were observed in general parenting behaviors taught by PAT suggesting adding SPHSK did not interfere with PAT efficacy as originally designed. Results indicate adding SPHSK to existing parent training can significantly enhance parents' awareness of and readiness to engage in protective behavioral strategies. Implementing SPHSK as a selective prevention strategy with at-risk parents receiving parent training through child welfare infrastructures is discussed.

Stress and information processing: acute psychosocial stress affects levels of mental abstraction

BACKGROUND

One mechanism by which acute psychosocial stress effects health-related cognitions and behaviors is through changes in the level of mental abstraction when processing information. However, it is unclear whether levels of mental abstraction would be higher or lower after an acute psychosocial stressor.

OBJECTIVES

This research examined acute psychosocial stress's impact on levels of mental abstraction.

DESIGN

Randomized between-subjects experimental design.

METHODS

A diverse sample of 164 undergraduate students aged 18-24 years old were randomly assigned to an acute psychosocial stressor or non-stressful control condition. Blood pressure (BP), heart rate (HR), and negative affect were monitored throughout the study and mental abstraction was measured at the end of each condition.

RESULTS

Mental abstraction was statistically significantly higher (i.e., more abstract) after the stress condition than after the control condition (p = 0.005, d = 0.44). This association was partially explained by negative affect (p = 0.017), but not BP or HR (ps > 0.60).

CONCLUSIONS

Acute psychosocial stress is associated with higher levels of mental abstraction after the stressor. These findings may have implications for stress-relevant interventions as accounting for the level of mental abstraction may enhance the efficacy of the intervention.

Assessment of health surveys: fitting a multidimensional graded response model

The multidimensional graded response model, an item response theory (IRT) model, can be used to improve the assessment of surveys, even when sample sizes are restricted. Typically, health-based survey development utilizes classical statistical techniques (e.g. reliability and factor analysis). In a review of four prominent journals within the field of Health Psychology, we found that IRT-based models were used in less than 10% of the studies examining scale development or assessment. However, implementing IRT-based methods can provide more details about individual survey items, which is useful when determining the final item content of surveys. An example using a quality of life survey for Cushing's syndrome (CushingQoL) highlights the main components for implementing the multidimensional graded response model. Patients with Cushing's syndrome (n = 397) completed the CushingQoL. Results from the multidimensional graded response model supported a 2-subscale scoring process for the survey. All items were deemed as worthy contributors to the survey. The graded response model can accommodate unidimensional or multidimensional scales, be used with relatively lower sample sizes, and is implemented in free software (example code provided in online Appendix). Use of this model can help to improve the quality of health-based scales being developed within the Health Sciences.

Associations of Parental General Monitoring with Adolescent Weight-Related Behaviors and Weight Status

OBJECTIVE

This study examined how parental general monitoring (PGM), which refers to parental awareness of adolescents' activities in various domains, is associated with adolescents' weight status and related behaviors and whether these relationships differ among racial/ethnic groups.

METHODS

Data are from 4,088 Black, Latino, and White youth assessed in seventh grade (mean age = 13.06). BMI percentile based on measured height and weight indicated weight status. PGM was assessed by adolescent report of parents' awareness of money spending, friends, and whereabouts. Adolescents reported both healthy and unhealthy dietary intake (DI), physical activity, and screen time over the past 7 days. Total sample and multigroup structural models were estimated.

RESULTS

PGM was associated with lower weight status for the total sample, as well as being positively associated with healthy DI and physical activity and inversely associated with screen time across racial/ethnic groups. PGM was also inversely associated with unhealthy DI, except for in Black adolescents. There was an indirect effect from PGM to lower weight status via reduced screen time among Latino and White adolescents.

CONCLUSIONS

Positive outcomes associated with PGM are extended to weight status and related behaviors. Efforts at improving weight status and related behaviors could benefit from addressing general parenting practices with their children.

The performance of the IES-R for Latinos and non-Latinos: Assessing measurement invariance

Violent acts on university campuses are becoming more frequent. Enrollment rates of Latinos at universities is increasing. Research has indicated that youths are more susceptible to trauma, particularly Latinos. Thus, it is imperative to evaluate the validity of commonly used posttraumatic stress measures among Latino college students. The Impact of Event Scale-Revised (IES-R) is one of the most commonly used metrics of posttraumatic stress disorder symptomatology. However, it is largely unknown if the IES-R is measuring the same construct across different sub-samples (e.g. Latino versus non-Latino). The current study aimed to assess measurement invariance for the IES-R between Latino and non-Latino participants. A total of 545 participants completed the IES-R. One- and three-factor scoring solutions were compared using confirmatory factor analyses. Measurement invariance was then evaluated by estimating several multiple-group confirmatory factor analytic models. Four models with an increasing degree of invariance across groups were compared. A significant χ2 difference test was used to indicate a significant change in model fit between nested models within the measurement invariance testing process. The three-factor scoring solution could not be used for the measurement invariance process because the subscale correlations were too high for estimation (rs 0.92-1.00). Therefore, the one-factor model was used for the invariance testing process. Invariance was met for each level of invariance: configural, metric, scalar, and strict. All measurement invariance testing results indicated that the one-factor solution for the IES-R was equivalent for the Latino and non-Latino participants.

Latent Growth Curve Models for Biomarkers of the Stress Response

Objective: The stress response is a dynamic process that can be characterized by predictable biochemical and psychological changes. Biomarkers of the stress response are typically measured over time and require statistical methods that can model change over time. One flexible method of evaluating change over time is the latent growth curve model (LGCM). However, stress researchers seldom use the LGCM when studying biomarkers, despite their benefits. Stress researchers may be unaware of how these methods can be useful. Therefore, the purpose of this paper is to provide an overview of LGCMs in the context of stress research. We specifically highlight the unique benefits of using these approaches. Methods: Hypothetical examples are used to describe four forms of the LGCM. Results: The following four specifications of the LGCM are described: basic LGCM, latent growth mixture model, piecewise LGCM, and LGCM for two parallel processes. The specifications of the LGCM are discussed in the context of the Trier Social Stress Test. Beyond the discussion of the four models, we present issues of modeling nonlinear patterns of change, assessing model fit, and linking specific research questions regarding biomarker research using different statistical models. Conclusions: The final sections of the paper discuss statistical software packages and more advanced modeling capabilities of LGCMs. The online Appendix contains example code with annotation from two statistical programs for the LCGM.

Using Person Fit Statistics to Detect Outliers in Survey Research

Context: When working with health-related questionnaires, outlier detection is important. However, traditional methods of outlier detection (e.g., boxplots) can miss participants with “atypical” responses to the questions that otherwise have similar total (subscale) scores. In addition to detecting outliers, it can be of clinical importance to determine the reason for the outlier status or “atypical” response.

Objective: The aim of the current study was to illustrate how to derive person fit statistics for outlier detection through a statistical method examining person fit with a health-based questionnaire.

Design and Participants: Patients treated for Cushing's syndrome (n = 394) were recruited from the Cushing's Support and Research Foundation's (CSRF) listserv and Facebook page.

Main Outcome Measure: Patients were directed to an online survey containing the CushingQoL (English version). A two-dimensional graded response model was estimated, and person fit statistics were generated using the Zh statistic.

Results: Conventional outlier detections methods revealed no outliers reflecting extreme scores on the subscales of the CushingQoL. However, person fit statistics identified 18 patients with “atypical” response patterns, which would have been otherwise missed (Zh > |±2.00|).

Conclusion: While the conventional methods of outlier detection indicated no outliers, person fit statistics identified several patients with “atypical” response patterns who otherwise appeared average. Person fit statistics allow researchers to delve further into the underlying problems experienced by these “atypical” patients treated for Cushing's syndrome. Annotated code is provided to aid other researchers in using this method.

Using subscales when scoring the Cushing's quality of life questionnaire

CONTEXT

Patients in long-term remission of Cushing's syndrome (CS) commonly report impaired quality of life (QoL). The CushingQoL questionnaire is a disease-specific QoL questionnaire for patients diagnosed with CS. The developers of the CushingQoL recommend using a global (total) score to assess QoL. However, the global score does not capture all aspects of QoL as outlined by the World Health Organization (WHO).

OBJECTIVE

The aim of the study was to compare the performance of different scoring options to determine the optimal method for the CushingQoL.

DESIGN AND PATIENTS

Patients in remission from CS (n=341) were recruited from the Cushing's Syndrome Research Foundation's email listserv and Facebook page, and asked to complete the CushingQoL and a short demographics survey.

RESULTS

Using an exploratory analysis, adequate model fit was obtained for the global score, as well as a 2-subscale (psychosocial issues and physical problems) scoring solution. Confirmatory methods were performed to identify the optimal scoring solution. Both the global score and the 2-subscale scoring solution showed adequate model fit. However, a χ(2) difference test indicated that the 2-subscale scoring solution was a significantly better fit than the global score (P<0.05).

CONCLUSION

If doctors or researchers would like to tease apart physical and psychosocial issues, the 2-subscale scoring solution would be recommended, since this solution showed to be optimal in scoring the CushingQoL. Regardless of the scoring solution used, the CushingQoL has proven to be a valuable resource for assessing health-related QoL in patients with CS.

Total score or subscales in scoring the acromegaly quality of life questionnaire: using novel confirmatory methods to compare scoring options

CONTEXT

Impaired quality of life (QoL) is common in patients after long-term remission of acromegaly. The acromegaly QoL (AcroQoL) is a disease-specific QoL questionnaire for patients diagnosed with acromegaly. The summed total score is the most frequently used scoring method of the AcroQoL. However, the total score does not capture all of the aspects of QoL that are outlined by the World Health Organization (WHO).

OBJECTIVE

The aim of the present study was to use novel and sophisticated confirmatory methods to identify the optimal number of subscales for the AcroQoL.

DESIGN AND PATIENTS

Patients in remission from acromegaly were recruited from the Leiden University Medical Center and were asked to complete the AcroQoL (Dutch version) questionnaire (n=72).

RESULTS

The three-subscale version of the AcroQoL consisted of subscales reflecting Physical Complaints, Appearance Issues, and Personal Relations Issues related to QoL. Model fit indices (i.e., comparative fit index and root mean square error of approximation) indicated that the three-subscale version represented the data better than the total score and two-subscale models did. A χ(2) difference test indicated that the three-subscale model was a significantly better fit than the total score and two-subscale models were (P<0.05).

CONCLUSION

Model fit and comparison statistics indicate that the three-subscale model is a better scoring method than the total score and two-subscale versions of the AcroQoL are. The three-subscale version also better reflected the WHO's recommendation of using a multidimensional measure of QoL than the total score and two-subscale methods did. Therefore, it is recommended that values from the three-subscales of the AcroQoL be reported in future research.