Susan L. Trumbetta

Trauma, PTSD, and the course of severe mental illness: an interactive model

Traumatic life events, as defined by DSM-IV, are common among persons with severe mental illnesses (SMI) such as schizophrenia. Limited evidence suggests concomitantly high rates of posttraumatic stress disorder (PTSD) in this population. However, conceptual models do not exist for understanding the interactions between trauma, PTSD, and SMI. We propose a model, which is an extension of the stress-vulnerability model, in which PTSD is hypothesized to mediate the negative effects of trauma on the course of SMI. Our model posits that PTSD influences psychiatric disorders both directly, through the effects of specific PTSD symptoms including avoidance, overarousal, and re-experiencing the trauma, and indirectly, through the effects of common correlates of PTSD such as retraumatization, substance abuse, and difficulties with interpersonal relationships. We discuss the evidence supporting this model, and consider several intervening variables that are hypothesized to moderate the proposed relationships between PTSD and SMI, including social support, coping and competence, and antisocial personality disorder. Theoretical and clinical implications of the model are considered, as well as several methodological and nosological issues. We conclude with a brief discussion of directions for future research aimed at evaluating components of the model.

Social networks and clinical outcomes of dually diagnosed homeless persons

Longitudinal associations between social network characteristics and clinical outcomes were examined in a group of 130 homeless persons diagnosed with substance abuse and severe mental illness. Patients whose alcohol-use disorder remitted over 18 months had more frequent contact with Alcoholics Anonymous (AA) or Narcotics Anonymous (NA) members at baseline, and those whose other substance use disorder remitted had fewer substance users in their baseline social networks. Smaller social networks predicted heavier alcohol use over time which, in turn, predicted attrition of abstinence supporters and decreased multiplexity of relationships. More frequent social contact predicted improved Expanded Brief Psychiatric Rating Scale (BPRS) symptoms. Improved BPRS symptoms and greater multiplexity of social relationships predicted each other over time. More severe BPRS symptoms and attrition of abstinence supporters predicted each other, but the mechanisms of this relationship and possible associations with alcohol or drug use remain to be tested. More severe BPRS symptoms also predicted later, proportional loss of professional contacts. Addiction Severity Index (ASI) drug use showed no longitudinal associations with social network characteristics.

The Washington University Twin Study of alcoholism

Genetic contributions to the liability to develop alcoholism in males of Northern and Western European ancestry are well‐established. However, questions remain concerning the role of genetic variation in the etiology of alcoholism among non‐white populations, among women, and the possibility of etiological heterogeneity in subtypes of alcoholism. The answers to these questions are needed to help define phenotypes for molecular genetic studies searching for QTLs for alcoholism. Twins from 295 pairs were consecutively ascertained at inpatient and outpatient psychiatric and alcohol treatment facilities in St. Louis, MO in 1981–1986. Probands and willing cotwins were evaluated by structured psychiatric interviews, psychometric assessment, and lifetime treatment records. One hundred fifty‐four probands met criteria for alcohol abuse/dependence (AAD), including twins from 45 MZ, 50 same‐sex DZ, and 59 opposite‐sex pairs. Twin‐pair resemblance was evaluated for AAD and alcohol dependence (AD), as well as for subsets defined by gender, patterns of comorbidity, ethnic background, and clinical features. Among males, heritability of AAD and AD was substantial, with little evidence for common environmental contributions to family resemblance. Pair resemblance among females was also substantial, but similar for MZ and DZ pairs, yielding near‐zero heritability estimates. However, based on these sample sizes, the sex differences were not statistically significant. The results confirm prior studies of strong genetic influences on alcoholism in males, but suggest lower genetic influence in females. Power to test other sources of heterogeneity was limited, but the results suggest no evidence for higher heritability for male early onset alcoholism or for alcoholism with comorbid antisocial personality.

Crime scene analysis and the escalation of violence in serial rape.

The current study examines the crime scene behavior manifest by 108 serial rapists responsible for the perpetration of 565 rapes across various cities within the US. The goal of the current study is to identify which aspects of crime scene behavior reported to law enforcement by the victim are most useful in predicting, early in a series of offenses, which rapists are most likely to escalate into higher and, at times, life threatening levels of violence. Using 58 scales that quantify the verbal, physical, and sexual behavior manifest by a rapist in his interaction with his victim during his first reported rape and 36 modal variables that summarized approach, timing, demographics, and weapon usage across the series of rapes, the study attempts to differentiate between those rapists who escalate in their use of blunt force (Increasers) from those who do not (Non-Increasers). A logistic regression indicates that rapists who are white rather than of minority status and who, at the time of their first reported rape, rape their victims for longer periods of time and use more profanity are more likely to escalate in their level of blunt force than those rapists who do not exhibit these behaviors. The relevance of this type of predictive framework for law enforcement in its attempts to prioritize particular investigations is discussed.

Changes in Adolescent Response Patterns on the MMPI/MMPI-A Across Four Decades

The purpose of this study was to explore changes in adolescent self-presentation on the Minnesota Multiphasic Personality Inventory (MMPI; Hathaway & McKinley, 1940) and MMPI-A (Butcher et al., 1992) over a 40-year period. The primary samples used for comparison in this study include 1,235 adolescents, age 14 through 16, derived from the MMPI-A normative sample (Butcher et al., 1992) collected in 1989 and 10,514 adolescents, age 14 through 16, collected in 1948 and 1954 from Hathaway and Monachesis (1963) study of adolescent personality and behavior. MMPI basic scale and item-level data were also included for 817 adolescents, age 14 through 16, collected by Colligan and Offord (1992) in 1985 as a further comparison. Between-sample analyses at the profile level revealed that adolescents from the MMPI-A normative sample scored significantly higher across basic clinical scales and lower on validity scales L and K than adolescents from the Hathaway and Monachesi (1963) sample, with mean data from the Colligan and Offord (1992) sample typically falling at a midpoint value. Analyses of Harris-Lingoes (Harris & Lingoes, 1955) subscale and item-level data were conducted to provide refined definitions of the contents of scale-level changes. Results were interpreted as reflecting moderate to large changes in response frequencies between eras of data collection, and emphasis was placed on the relatively high frequency of item endorsements by contemporary adolescents in the clinical direction in the MMPI-A normative sample. A series of cautions and limitations are also offered in interpreting these patterns.

The etiology of mathematical and reading (dis)ability covariation in a sample of Dutch twins

The genetic etiology of mathematical and reading (dis)ability has been studied in a number of distinct samples, but the true nature of the relationship between the two remains unclear. Data from the Netherlands Twin Register was used to determine the etiology of the relationship between mathematical and reading (dis)ability in adolescent twins. Ratings of mathematical and reading problems were obtained from parents of over 1500 twin pairs. Results of bivariate structural equation modeling showed a genetic correlation around .60, which explained over 90% of the phenotypic correlation between mathematical and reading ability. The genetic model was the same for males and females.

Endophenotypes for Marital Status in the NAS-NRC Twin Registry

This study explored two possible endophenotypes for marital status: 1) the predisposition to form and maintain lasting pair bonds and 2) the predisposition to have multiple mates over the life span. These endophenotypes were constructed using 1972 and 1985 marital status data from a followed-up subsample of the NAS-NRC WWII Veteran Twin Registry. In the 1972 data, consisting of 2297 MZ and 2443 DZ twin pairs, 42% of the variance in pair bonding could be attributed to additive genetic and 58% to nonshared environmental factors and measurement error. Of the variance in multiple mates, 28% could be attributed to additive genetic and 62% to nonshared environment/error factors. In the 1985 data, consisting of 1359 MZ and 1208 DZ twin pairs, 31% of the variance in pair bonding could be attributed to non-additive genetic and 69% to nonshared environment/error factors. Of the variance in multiple mates, 22% could be attributed to additive genetic and 78% to nonshared environment/error factors. Although parameter estimates were marked by wide confidence intervals, no variance in either endophenotype could be attributed to common family environment.

A Danish Twin Study of Schizophrenia Liability: Investigation from Interviewed Twins for Genetic Links to Affective Psychoses and for Cross-Cohort Comparisons.

We studied schizophrenia liability in a Danish population-based sample of 44 twin pairs (13 MZ, 31 DZ, SS plus OS) in order to replicate previous twin study findings using contemporary diagnostic criteria, to examine genetic liability shared between schizophrenia and other disorders, and to explore whether variance in schizophrenia liability attributable to environmental factors may have decreased with successive cohorts exposed to improvements in public health. ICD-10 diagnoses were determined by clinical interview. Although the best-fitting, most parsimonious biometric model of schizophrenia liability specified variance attributable to additive genetic and non-shared environmental factors, this model did not differ significantly from a model that also included non-additive genetic factors, consistent with recent interview-based twin studies. Schizophrenia showed strong genetic links to other psychotic disorders but much less so for the broader category of psychiatric disorders in general. We also observed a marginally significant decline in schizophrenia variance attributable to environmental factors over successive Western European cohorts, consistent perhaps with improvements in diagnosis and in prenatal and perinatal care and with a secular decline in the prevalence of schizophrenia in that region.

Crime scene analysis and the escalation of violence in serial rape

The current study examines the crime scene behavior manifest by 108 serial rapists responsible for the perpetration of 565 rapes across various cities within the US. The goal of the current study is to identify which aspects of crime scene behavior reported to law enforcement by the victim are most useful in predicting, early in a series of offenses, which rapists are most likely to escalate into higher and, at times, life threatening levels of violence. Using 58 scales that quantify the verbal, physical, and sexual behavior manifest by a rapist in his interaction with his victim during his first reported rape and 36 modal variables that summarized approach, timing, demographics, and weapon usage across the series of rapes, the study attempts to differentiate between those rapists who escalate in their use of blunt force (Increasers) from those who do not (Non-Increasers). A logistic regression indicates that rapists who are white rather than of minority status and who, at the time of their first reported rape, rape their victims for longer periods of time and use more profanity are more likely to escalate in their level of blunt force than those rapists who do not exhibit these behaviors. The relevance of this type of predictive framework for law enforcement in its attempts to prioritize particular investigations is discussed.