Validation of GAD-7 for postpartum anxiety screening in Latvia: psychometric and diagnostic performance

Introduction

Postpartum anxiety [ER1] is increasingly recognised as a clinically significant condition, which can affect the whole family’s wellbeing, especially mother and infant outcomes [1]. It remains undetected in routine care. Brief screening tools, such as the Generalised Anxiety Disorder 7 scale (GAD-7), are proven to facilitate early identification of anxiety symptoms in the postpartum population [2]. However, its validity, particularly in smaller European countries with distinct demographic profiles such as Latvia, remains insufficiently studied.

Aims

To examine the internal consistency and diagnostic accuracy of the GAD-7 in a Latvian postpartum sample, and to identify an empirically grounded screening cut-off point against a structured diagnostic interview.

Methods

This cross-sectional study was conducted among women attending routine gynaecological visits 4–6 weeks postpartum at Riga Maternity Hospital over a 12-month period. Participants were screened with the GAD-7 and were eligible if they scored ≥5. The Mini International Neuropsychiatric Interview, version 7.0.2 (MINI), served as the diagnostic standard and was administered remotely by a trained psychiatrist within 2 weeks of screening. Internal consistency was evaluated using Cronbach’s α. Diagnostic performance was assessed by receiver operating characteristic (ROC) analysis, including sensitivity, specificity, positive and negative predictive values (PPV, NPV), likelihood ratios (LR+, LR–), and Youden Index (J) across various cut-off scores.

Results

In total, 272 women completed the GAD-7 screening, of whom 109 (40.07%) scored ≥5 and were classified as screening positive (mean age 30.66±5.54 years). Of these, 92 women completed the MINI and were included in the diagnostic accuracy analysis. Forty-two participants (45.65%) met the diagnostic criteria for Generalised Anxiety Disorder (GAD) according to the MINI. The Cronbach’s α for all GAD-7 items was 0.84. The ROC analysis (n=92) yielded an area under the curve of 0.644 (95% CI: 0.531 - 0.756; p < 0.018). At a cut-off score of ≥6, the GAD-7 demonstrated the best overall performance, with a sensitivity of 0.83, specificity of 0.40, the highest J of 0.23, and an LR+ of 1.39. Results across cut-off scores are shown in the table below. At higher thresholds (e.g. ≥10), specificity increased while sensitivity declined.

Cut-Off (GAD-7)	Sensitivity	Specificity	PPV	NPV	LR+	LR-	J
≥ 1	1.000	0.020	0.462	1.000	1.020	0.000	0.020
≥ 2	1.000	0.020	0.462	1.000	1.020	0.000	0.020
≥ 3	0.952	0.100	0.471	0.714	1.058	0.476	0.052
≥ 4	0.952	0.160	0.488	0.800	1.134	0.298	0.112
≥ 5	0.881	0.280	0.507	0.737	1.224	0.425	0.161
≥ 6	0.833	0.400	0.539	0.741	1.389	0.417	0.233
≥ 7	0.691	0.500	0.537	0.658	1.381	0.619	0.191
≥ 8	0.548	0.680	0.590	0.642	1.711	0.665	0.228
≥ 9	0.429	0.720	0.563	0.600	1.531	0.794	0.149
≥ 10	0.381	0.780	0.593	0.600	1.732	0.794	0.161

Conclusions

The GAD-7 showed good internal consistency and moderate diagnostic accuracy for detecting GAD in postpartum women in Latvia when validated against the MINI. A cut-off score of ≥6 provided the best balance between sensitivity and specificity, supporting its use as a screening tool. These findings indicate that the GAD-7 is a useful first-step instrument for postpartum anxiety screening in Latvia.

References

1. E. Dowse et al., “Impact of Perinatal Depression and Anxiety on Birth Outcomes: A Retrospective Data Analysis,” Matern. Child Health J., vol. 24, no. 6, pp. 718–726, Jun. 2020, doi: 10.1007/s10995-020-02906-6.
2. A. Stefana, F. Mirabella, A. Gigantesco, and L. Camoni, “Factor structure, measurement invariance, and clinical change benchmarks of the Generalized Anxiety Disorder-7 (GAD-7) in pregnancy and postpartum,” J. Anxiety Disord., vol. 118, p. 103123, Mar. 2026, doi: 10.1016/J.JANXDIS.2026.103123.