Prediction of risk of depressive symptoms in menopausal women based on hot flash and sweating symptoms: a multicentre study

Abstract

Objective: The present study aimed to develop a symptom-based (namely, hot flashes and sweating) scoring system for predicting the risk of depressive symptoms in menopausal women via a multicentre cross-sectional survey.

Methods: The data examined in the present study were obtained from 1,004 women aged 40-60 years who underwent physical examination at A Hospital. The basic information was obtained using a questionnaire-based survey. A self-rating depression scale was used to obtain the depressive symptom scores, while the Kupperman Menopausal Index was used to obtain the scores for the frequency of hot flashes and sweating. A logistic regression model was also established. The resulting β coefficient was employed to calculate and predict the risk of depressive symptoms in these women and a risk scoring system was established. The scoring system was validated using samples from 2 other centers (validation sample 1: B Hospital, 440 women; validation sample 2: C Hospital, 247 women).

Results: The scoring system developed to predict the risk of depressive symptoms in menopausal women was based on hot flash and sweating symptoms and associated with menopausal status, hot flash scores, education level (high school education and below) and being diabetic. The scoring system yielded a total score of 0-54 points. For women in the study sample, the area under the curve (AUC) of depressive symptom risk score was 0.750 (95% CI, 0.708-0.793). Validation sample 1 had an AUC of 0.731 (95% CI, 0.667-0.794), while validation sample 2 had an AUC of 0.744 (95% CI, 0.669-0.820). The optimal cut-off score to assess depressive symptoms in women participating in the present study was 31 points. The sensitivity and specificity for predicting depressive symptoms in the study sample were 0.667 and 0.701, respectively. In contrast, the sensitivity was 0.840 in validation sample 1 and 0.879 in validation sample 2.

Conclusion: The hot flash and sweating symptom-based scoring system developed to predict the risk of depressive symptoms in menopausal women relies on non-laboratory survey data. The system is simple, practical, and convenient to use. For Chinese huge population of menopausal women, the scoring system should be considered a reliable screening tool for depressive symptoms.

Keywords: depressive symptoms; hot flash; menopause; sweating.

Figure 1

Process flow diagram. Abbreviation: SDS, self-rating depression scale.

Figure 2

ROC curves for depressive symptom risk scores. Notes: Risk scores for prediction of the occurrence of depressive symptoms. (A) Study sample; (B) validation sample 1; (C) validation sample 2. Abbreviation: ROC, receiver operating characteristic.