Psychometric properties of stress and anxiety measures among nulliparous women

Abstract

Objective: To examine the psychometric properties of three measures, the perceived stress scale (PSS), pregnancy experience scale (PES), and state trait anxiety inventory (STAI), for assessing stress and anxiety during pregnancy among a large sample of nulliparous women.

Methods: The sample included 10,002 pregnant women participating in the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be (nMoM2b). Internal consistency reliability was assessed with Cronbach’s alpha and factorial validity with confirmatory factor analyses. Intraclass correlations (ICCs) were calculated to determine stability of PSS scales over time. Psychometric properties were examined for the overall sample, as well as subgroups based on maternal age, race/ethnicity and language.

Results: All three scales demonstrated good internal consistency reliability. Confirmatory factor analyses supported the factor structures of the PSS and the PES. However, a one-factor solution of the trait-anxiety subscale from the STAI did not fit well; a two-factor solution, splitting the items into factors based on direction of item wording (positive versus negative) provided a better fit. Scores on the PSS were generally stable over time (ICC = 0.60). Subgroup analyses revealed a few items that did not perform well on Spanish versions of the scales.

Conclusion: Overall, the scales performed well, suggesting they could be useful tools for identifying women experiencing high levels of stress and anxiety during pregnancy and allowing for the implementation of interventions to help reduce maternal stress and anxiety.

Keywords:

• Anxiety
• parity
• psychometrics
• stress

Disclosure statement

The authors report no conflicts of interest.

Current knowledge on this subject

• Stress and anxiety during pregnancy have been associated with poor outcomes among mothers and their children.

• Trajectories of stress through the prenatal to postpartum periods have been shown to differ among women based on parity.

• Reliable and valid measures are needed to identify pregnant women experiencing high levels of stress and anxiety.

• The psychometric properties of most stress and anxiety measures have not been assessed among nulliparous women.

What this study adds

• Two stress-related measures (perceived stress scale and pregnancy experience scale) and one anxiety measure (state trait anxiety inventory – trait-anxiety subscale) demonstrated good psychometric properties among a large sample of nulliparous women.

• The measures generally performed well across age and racial/ethnic subgroups.

• Scores on the perceived stress scale were positively correlated between 8–13 and 22–29 weeks’ gestation.