Oral contraceptive use is not related to gender self-concept

Abstract

Oral contraceptive (OC) use is a valuable and underutilized natural experiment for examining exogenous hormone effects on aspects of cognition and behavior, especially aspects that show sex or gender differences. OC use, however, is not randomly assigned; women who use OCs may systematically differ from non-users for gender-linked reasons that potentially confound investigations of the pharmacokinetic effects of OCs. Thus, the primary aim of this study was to investigate whether OC users and non-users differ in gender self-concepts, as sociocultural factors might influence who uses OCs and why. A secondary aim was to examine whether OC users of different pill formulations differ in gender self-concept. Multivariate analyses of covariance (controlling for age) on a large sample of 395 young adult women revealed no significant differences between OC users (n = 247) and non-users (n = 148) in four aspects of gender self-concept (expressivity, instrumentality, femininity, and masculinity), with univariate effect sizes averaging d = 0.07. Moreover, inferences about group differences did not change when considering sociocultural factors (i.e., reason for OC use) or pharmacokinetic properties of the pills (i.e., androgenicity). Thus, findings indicate that differences in gender self-concept do not underlie OC use and non-use, and that the androgenicity of different OC formulations does not impact gender self-concept. Taken together, findings also highlight the specificity of exogenous ovarian hormone influences, as there are documented effects on cognition but not on personal characteristics, such as gender self-concept.

Keywords: Androgens; Expressivity; Femininity; Hormonal contraceptives; Instrumentality; Masculinity; Ovarian hormones.

Figure 1.

Comparisons on gender self-concepts of personality attributes (expressivity and instrumentality) and self-categorizations (femininity and masculinity) between: (1A) oral contraceptive (OC) users and non-users (Non); (1B) non-users (duplicated from A for clarity) and OC users with different reasons; and (1C) OC users of pills with different androgenicity. Stacked bars show group means; error bars show standard deviations; Cohen’s d reflects pairwise comparisons. There were no significant differences.