INTERNALIZING SYMPTOM SCALE BASED ON THE ADOLESCENT AND ADULT SSAGA IN A PROSPECTIVE STUDY OF FAMILIES AT HIGH RISK FOR ALCOHOLISM

L ACION; J KRAMER; X LIU; G CHAN; K BUCHOLZ; V HESSELBROCK; J VAIDYA; V MCCUTCHEON

Conferencia; ANNUAL RSA SCIENTIFIC MEETING; 2016

Overall internalizing (INT) characteristics show stronger associations with problematic alcohol usethan do individual INT disorders (e.g., anxiety, depression). The Semi-Structured Assessment forthe Genetics of Alcoholism(SSAGA) is a widely used interview allowing reliable establishment of lifetime DSM-IV psychiatric diagnoses. Our goal was to develop an INT symptom scale based on the SSAGA. We developed a 7-item scale (total score range: 0?14) by extracting lifetime DSM-IV-based symptomatic and diagnostic information about social phobia, panic disorder, agoraphobia, obsessive compulsive disorder, post-traumatic stress disorder, major depressive disorder, and suicidality. Higher scores indicate more INT characteristics.We used two samples of 1,646 children and 1,059 young adults. Both were from high-risk families in a multi-site alcoholismstudy.We evaluated internal consistency using ordinal a, an analogous to Cronbach?s a based on polychoric correlation.We assessed construct validity by examining the association between the proposed INT scale and ageappropriate versions of the Achenbach Self Report (ASR). Bivariate associations between the INT scale score (ISS) and sex, age, and alcohol characteristics were also examined. The INT scale exhibited very good internal consistency (a = 0.82 for both samples). Positive associations between the ISS and the INT Problems raw score of the ASR supported construct validity. The ISS was higher for females than formales and increased with age in both samples. In young adults, there was a positive association between the ISS and the number of alcohol abuse or dependence symptoms.We also found a positive association between ISS and themaximumnumber of drinks in 24 h and a negative association of the ISS with age of first drink in young adults. Young adults who ever had a blackout had higher ISS than those who did not. All findings were significant with p-values < 0.0001. When we examined symptom counts of individual INT disorders (i.e.,major depressive disorder, post-traumatic stress disorder, and panic disorder), all associations with alcohol characteristics were weaker than those with the ISS, and statistical significance was lost in some cases. In sum, our INT scale extracts all relevant information fromthe SSAGA, has very good psychometric properties, can be used to study the association between INT characteristics and problematic alcohol use, and may be a more robustmeasure than individual INT disorders.