Drug use and intimate relationships among women and men: separating specific from general effects in prospective data using structural equation models.

This article presents the use of specific paths in structural equation models (SEMs) that allow separation and tests of both unique and common (latent) aspects of measured variables. Cross-sectional studies have established that problems in intimate relationships are related to drug use, but the direction of these effects is unclear and is best addressed by using SEMs with prospective data. Reciprocal effects between constructs of polydrug use and relationship quality were examined for 307 young women and 117 young men over a 4-year period while social conformity was controlled for. No cross-lagged effects were found between latent constructs, except for polydrug use, which increased later divorces for women. However, specific paths allowed tests of types of drug use and relationship quality as both predictors and consequences. There were 8 effects for women and 6 for men from specific measures of drug use that adversely affected aspects of relationship quality. Conversely, only one effect was found from relationship quality to drug use variables for both genders.