Scale validity of the PHQ 8 and GAD 7 as severity measures in young adults in Sweden

Background

The Patient Health Questionnaire‑8 (PHQ‑8) and the Generalized Anxiety Disorder‑7 (GAD‑7) are widely used measures of depressive and anxiety symptoms in population‑based surveys. Validation has mainly focused on diagnostic case‑finding, with less attention to whether total scores meet the measurement requirements necessary to function as severity scales.

Methods

Data were drawn from Futura01, a Swedish longitudinal population survey study following a school‑based cohort from adolescence into adulthood. The present study used survey round collected in 2024, when participants were approximately 23 years old (n = 3365). Scale properties of the PHQ‑8 and GAD‑7 were examined using Mokken scale analysis. Scalability was evaluated using Loevinger’s H coefficients, and monotonicity was assessed to support ordinal interpretation of item responses. Convergent validity was examined by relating symptom severity levels to functional impairment (the Work and Social Adjustment Scale, WSAS).

Results

PHQ-8 and GAD-7 both demonstrated clear hierarchical structure and satisfied the assumption of monotonicity. The PHQ‑8 showed moderate to strong scalability (H = 0.49), with item scalability coefficients (Hi) ranging from 0.43 to 0.55. The GAD‑7 showed strong scalability (H = 0.58), with item coefficients ranging from 0.45 to 0.67. No violations of monotonicity were identified for either scale. Functional impairment increased monotonically across increasing levels of depressive and anxiety symptom severity.

Key messages

• PHQ 8 and GAD 7 demonstrate scale validity as severity measures, allowing total scores to be interpreted as ordered levels of symptom severity rather than only tools for diagnostic case finding.

• This strengthens the use of PHQ-8 and GAD-7 as instruments for population mental health surveillance.