Sex disparities in substance abuse research: Evaluating 23 years of structural neuroimaging studies

Abstract

Background: Sex differences in brain structure and clinical course of substance use disorders underscores the need to include women in structural brain imaging studies. The NIH has supported the need for research to address sex differences. We evaluated female enrollment in substance abuse structural brain imaging research and the methods used to study sex differences in substance effects.

Methods: Structural brain imaging studies published through 2016 (n=230) were evaluated for number of participants by sex and substance use status and methods used to evaluate sex differences. Temporal trends in the numbers of participants by sex and substance use status were analyzed. We evaluated how often sex effects were appropriately analyzed and the proportion of studies that found sex by substance interactions on volumetric measures.

Results: Female enrollment increased over time, but remained significantly lower than male enrollment (p=0.01), with the greatest bias for alcohol and opiate studies. 79% of studies included both sexes; however, 74% did not evaluate sex effects or used an analytic approach that precluded detection of sex by substance use interactions. 85% of studies that stratified by sex reported different substance effects on brain volumes. Only 33% of studies examining two-way interactions found significant interactions, highlighting that many studies were underpowered to detect interactions.

Conclusions: Although female participation in substance use studies of brain morphometry has increased, sex disparity persists. Studying adequate numbers of both sexes and employing correct analytic approaches is critical for understanding sex differences in brain morphometric changes in substance abuse.

Keywords: Female inclusion; Grey matter volume; Sexual dimorphism; Structural brain imaging; Substance abuse.

Figure 1

Article inclusion and exclusion (PRISMA Diagram).

Figure 2

Evaluation Results of Statistical Methodology.

Figure 3. Enrollment by sex and substance use status over time

Top, number of NIH studies and all studies meeting inclusion criteria each year. Second box from top, total number of participants in the literature each year by group and sex. Third box from top, mean number of participants per study per year by group and sex. Bottom box, mean number of participants in NIH-funded studies per study per year by group and sex. Male and female enrollment were significantly different in SUD (p=0.01) and controls (p=0.02), with systematically higher enrollment of male participants in both groups over time.

Figure 4. Female to male sampling ratio by substance type

For each study in the sample population, the ratio of total females to males was calculated to form a distribution with cardinality equal to the number of included studies. The top and bottom edges of each box represent the 3^rd and 1^st quartiles, and lines within the boxes indicate medians. Whiskers above and below each box represent the 90^th and 10^th percentiles. Statistical outliers are plotted as points.