The generalizability of empirically derived syndromes of collateral-reported elder psychopathology across 11 societies

The purpose of this study was to test whether a syndrome model of elder psychopathology derived from collateral ratings, such as from spouses and adult children, in the United States would be generalizable in 11 other societies. Societies represented South America, Asia, and Europe. The Older Adult Behavior Checklist (OABCL) was completed by collateral informants for 6141 60- to 102-year-olds. The tested model comprised syndromes designated as Anxious/Depressed, Worries, Somatic Complaints, Functional Impairment, Memory/Cognition Problems, Thought Problems, and Irritable/Disinhibited. The model was tested using confirmatory factor analyses in each society separately. The primary model fit index showed a good fit for all societies, while the secondary model fit indices showed acceptable to a good fit for all societies. The items loaded strongly on their respective factors, with a median item loading of 0.69 across the 11 societies. By syndrome, the overall median item loadings ranged from 0.47 for Worries to 0.77 for Functional Impairment. The OABCL syndrome structure was thus generalizable across the tested societies. The OABCL can be used for broad assessment of psychopathology for elders of diverse backgrounds in nursing services and research.

The generalizability of Older Adult Self-Report (OASR) syndromes of psychopathology across 20 societies

OBJECTIVES

As the world population ages, psychiatrists will increasingly need instruments for measuring constructs of psychopathology that are generalizable to diverse elders. The study tested whether syndromes of co-occurring problems derived from self-ratings of psychopathology by US elders would fit self-ratings by elders in 19 other societies.

METHODS/DESIGN

The Older Adult Self-Report (OASR) was completed by 12 826 adults who were 60 to 102 years old in 19 societies from North and South America, Asia, and Eastern, Northern, Southern, and Western Europe, plus the United States. Individual and multigroup confirmatory factor analyses (CFAs) tested the fit of the seven-syndrome OASR model, consisting of the Anxious/Depressed, Worries, Somatic Complaints, Functional Impairment, Memory/Cognition Problems, Thought Problems, and Irritable/Disinhibited syndromes.

RESULTS

In individual CFAs, the primary model fit index showed good fit for all societies, while the secondary model fit indices showed acceptable to good fit. The items loaded strongly on their respective factors, with a median item loading of .63 across 20 societies, and 98.7% of the loadings were statistically significant. In multigroup CFAs, 98% of items demonstrated approximate or full metric invariance. Fifteen percent of items demonstrated approximate or full scalar invariance, and another 59% demonstrated scalar invariance across more than half of societies.

CONCLUSIONS

The findings supported the generalizability of OASR syndromes across societies. The seven syndromes offer empirically based clinical constructs that are relevant for elders of different backgrounds. They can be used to assess diverse elders and as a taxonomic framework to facilitate communication, services, research, and training in geriatric psychiatry.

Syndromes of collateral-reported psychopathology for ages 18-59 in 18 Societies

The purpose was to advance research and clinical methodology for assessing psychopathology by testing the international generalizability of an 8-syndrome model derived from collateral ratings of adult behavioral, emotional, social, and thought problems. Collateral informants rated 8,582 18-59-year-old residents of 18 societies on the Adult Behavior Checklist (ABCL). Confirmatory factor analyses tested the fit of the 8-syndrome model to ratings from each society. The primary model fit index (Root Mean Square Error of Approximation) showed good model fit for all societies, while secondary indices (Tucker Lewis Index, Comparative Fit Index) showed acceptable to good fit for 17 societies. Factor loadings were robust across societies and items. Of the 5,007 estimated parameters, 4 (0.08%) were outside the admissible parameter space, but 95% confidence intervals included the admissible space, indicating that the 4 deviant parameters could be due to sampling fluctuations. The findings are consistent with previous evidence for the generalizability of the 8-syndrome model in self-ratings from 29 societies, and support the 8-syndrome model for operationalizing phenotypes of adult psychopathology from multi-informant ratings in diverse societies.