Background
Diabetes is recognized as one of the most challenging chronic diseases worldwide, with the global prevalence rising from 9% in 2019 to 10.5% in 2021 according to the International Diabetes Federation [
1]. People with diabetes not only face life-threatening symptoms and disabling complications but may also suffer from psychosocial problems. According to the International Prevalence and Treatment of Diabetes and Depression Study (INTERPRET-DD), which spans 14 countries, the prevalence of depression in patients with diabetes reached 10.6%, with 17% of patients experiencing moderate to severe depressive symptoms [
2]. In China, the combined prevalence of depression in patients with type 2 diabetes was 25.9%, ranging from 0.8 to 52.6%, according to different depression screening tools [
3].
Numerous shreds of evidence have proved that depression may adversely influence glycemic control, comorbidity, and even mortality of diabetes [
4‐
10]. Depression may also interrupt patients’ behavior and well-being, including diabetes self-care, diet control, physical exercise, and self-efficacy [
11‐
15]. On the contrary, managing the psychosocial problems in diabetes mellitus can promote glucose control, dietary behavior, and life quality [
16]. Various aspects of psychosocial well-being, such as self-efficacy, positive affect, optimism and resilience have indicated positive correlations with superior diabetes management and better outcomes [
17‐
19]. Under these circumstances, screening for depression and identifying well-being status are of great significance for diabetes mellitus patients. Thus, proper screening instruments are essential to help both endocrinologists and psychiatrists identify the depressive disorders in these patients to provide timely intervention.
Several indices have been developed to evaluate psychological well-being, among which the 5-Item WHO Well-Being Index (WHO-5) has displayed satisfactory sensitivity and validity measurement of subjective well-being [
20,
21]. All the item content of WHO-5 focused on positive well-being, such as “cheerful and in good spirits”, and “calm and relaxed”, which is rather different from other depression screening instruments. Thus, WHO-5 can be also used as a reliable screening tool for depression, as confirmed in various studies related to diabetes with good psychometric properties [
20,
22]. Meanwhile, the psychometric properties of the Chinese Version of WHO-5 (WHO-5-C) have also been validated in a sample of 1414 Chinese university students [
23]. However, it has not been implicated in a medical setting concerning diabetes. Given that Chinese are usually ashamed to express and admit bad mental feelings, WHO-5 is more acceptable culturally. On the other side, the WHO-5 is brief and efficient for assessment with only 5 items, thus measured conveniently in bustling Chinese clinics. The acceptance and effectiveness of the screening instrument should not be based merely on its psychometric properties, but take culture, language, and literacy into consideration.
Therefore, the purpose of this study was to explore the psychometric properties of the WHO-5-C in screening for depression in patients with type 2 diabetes.
Methods
Participants
The cross-sectional study was conducted to evaluate the reliability and validity of the Chinese Version of the 5-item WHO Well-Being Index in type 2 diabetes patients. All participants were recruited from the Department of Endocrinology in Peking Union Medical College Hospital from July 2014 to March 2015.
Inclusion criteria: 1) aged between 18 to 65 years old; 2) diagnosed with type 2 diabetes for at least 1 year; 3) outpatients; 4) could sign the informed consent independently.
Exclusion criteria: 1) diagnosed as type 2 diabetes within 1 year; 2) illiterate, cognitively impaired, or patients in serious medical condition (e.g., delusional, delirium, or acute suicidal tendency) that could not complete the survey; 3) hospitalized patients; 4) involved in other interventional studies within 3 months; 5) pregnant or within the six months of lactation.
Ethical approval was obtained from the institutional review board of Peking Union Medical College Hospital (registration number S-713) as part of the International Diabetes and Depression Comorbidity and Treatment research, and written informed consent was obtained from all the participants.
Procedure
Before completing the questionnaires, sociodemographic information such as sex, age, education level, marriage state, living site, and income level was collected from patients. Questionnaires were then provided to patients to finish, including the WHO-5-C, the 9-item Patient Health Questionnaire (PHQ-9), and the 20-item Problem Areas in Diabetes (PAID-20). After finishing the questionnaires, patients were then assessed by a research assistant who was unaware of the patients’ WHO-5-C scores and was trained in the use of the Mini International Neuropsychiatric Interview (M.I.N.I.), Hamilton Rating Scale for Depression (HAM-D).
Measures
The WHO-5 consists of five items with a 6-point Likert-type scale ranging from 0 (at no time) to 5 (all of the time) that measure well-being. The raw score ranging from 0 to 25 was then multiplied by 4 to give the final score. A higher score indicates a higher level of well-being [
20]. The WHO-5-C experienced translation and back-translation procedures by two individual translators and was further convinced in two pilot studies, reporting no difficulty in understanding the questions [
24]. Internal consistency was confirmed in several studies [
23].
The M.I.N.I. is a structured clinical diagnostic interview schedule standardized for the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition Axis-I Disorders [
25]. It can be carried out reliably by properly trained interviewers. The depression modules of the schedule were used in the study.
The HAM-D is a mature tool for depression severity evaluation and prognosis prediction. The Chinese version of the HAM-D has acceptable inter-item consistency (Cronbach’s α = 0.714) and concurrent validity, holding a negative correlation with the Global Assessment Scale (Spearman’s correlation coefficient = -0.487,
p < 0.001) [
26].
The Chinese version of PAID-20 measures emotional problems in diabetes patients by a 20-item self-report scale. A positive relationship with Hb
A1C was revealed was reported (Spearman’s correlation coefficient = 0.15,
p < 0.05), indicating that levels of diabetes-related distress were related to poor glycemic control [
27].
PHQ-9 has been developed as a valid screening tool for depression, broadly utilized in primary care settings. The Chinese version of PHQ-9 has been validated with Cronbach’s α ranging from 0.765 to 0.983, pooled sensitivity of 0.88 (95% CI 0.80–0.87), and pooled specificity of 0.87 (95%CI 0.83–0.91), proving its significance for clinical use [
28].
Statistical analysis
Statistical Package Social Sciences (SPSS) 25.0 statistical software and Mplus 8.8 and AMOS 28 were employed to perform statistical analyses.
SPSS was used to establish internal consistency reliability by Cronbach’s α value. Cronbach’s α values of 0.70 were regarded as acceptable and 0.90 as excellent [
29]. Mplus and AMOS were used to run a confirmatory factor analysis [estimation method = diagonal weighted least square] to test the factorial structure of the WHO-5-C in the Chinese diabetes samples. The Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), and the root mean square error of approximation (RMSEA) were used to test global model fits. A good fit for continuous data is indicated for Comparative Fit Index > 0.95 [
30], Tucker-Lewis Index > 0.95 [
31], and root mean square error of approximation < 0.08 [
32].
The concurrent validity of the WHO-5-C was analyzed by estimating the relationship of WHO-5-C with HAM-D, PHQ-9, PAID-20, and clinical indicators. The criterion validity of the WHO-5-C was tested using the relevant sections of the M.I.N.I. DSM-IV depression and PHQ-9. The receiver operating characteristic (ROC) curve was used to determine the level of accuracy with which the WHO-5-C can predict depression diagnosis.
Discussion
Our findings suggest good reliability and validity for the Chinese Version of WHO-5 and support its utility as a suitable screening instrument for likely depression among Chinese patients with type 2 diabetes. The WHO-5-C presents good internal consistency, with Cronbach’s alpha values of 0.85, similar to the values reported in other countries [
33‐
35] and other medical settings in China [
36,
37]. Construct validation was confirmed by CFA in our study. The results show WHO-5-C fits the unidimensional factor model well among diabetic patients, consistent with that of the original or other versions of WHO-5-C [
20].
We also determined the concurrent validity of the WHO-5-C in relation to the HAM-D and PHQ-9, both obtaining significant negative correlations (
r = -0.610 and -0.694, respectively), which indicates the comparative role of WHO-5-C as a screening instrument for depression in diabetes patients. This result is consistent with that of the original or other language versions of WHO-5 in diabetic patients [
29,
38‐
42]. The correlation coefficient of WHO-5-C with PAID is relatively low (
r = -0.466), but also significant (
p < 0.001). PAID evaluates diabetes-related distress specifically, an adverse psychological burden mainly on account of rigorous diabetes management and worries about complications. WHO-5 assesses depression, a persistent low mood, and a lack of interest and motivation. The psychological characteristics reflected in the two scales partly overlap, but are different, which may contribute to the above lower coefficient value. We failed to reveal a significant correlation between the WHO-5-C score and other laboratory indicators, such as glycosylated hemoglobin. However, Papanas and Prinz’s studies both confirmed the significant association between WHO-5 and glycemic control [
43,
44]. Our previous findings also suggest that depression and diabetes-related distress interact to affect glycemic control. Another cross-sectional study even indicated that the score of WHO-5 was reversely correlated with the number of diabetes complications [
45]. We assumed that the reason for this discrepancy may lie in that how the psychological state affects the physical indicators is still unclear. Personality traits, coping styles, and resource support may all play a role in it. A further longitudinal study is required to clarify the predictive validity of WHO-5 for glycemic control and diabetes complications.
The ROC analysis indicated that the Chinese version of WHO-5-C has sufficient power to screen major depression patients from all diabetes mellitus patients. The AUC was close to 1, indicating satisfactory accuracy for diabetes screening. We discovered that at a cut-off point of < 42, the WHO-5-C had the best sensitivity (83.1%) and specificity (75.0%). However, the cut-off value is different from the original version of WHO-5-C, which is < 52 [
20]. Most of the studies about other language versions of WHO-5 regarding diabetes also take the cut-off score of ≤ 50, reporting a sensitivity from 0.57 to 1.00 and specificity from 0.78 to 0.88 [
20]. Other studies that examined the utility of the WHO-5 English Version as a screening instrument also took < 52 as a cut-off score, which displayed adequate sensitivity and specificity [
46,
47]. We think the differences are mainly due to two reasons. The first is the index of validity. The most frequently chosen index of validities is CES-D ≥ 16, or PHQ-9 ≥ 10 in diabetes patients [
46‐
49]. However, we used M.I.N.I. DSM-IV depression as a golden standard in diagnosis, which may lead to a discrepancy in the cut-off value. Compared with other studies using M.I.N.I. as a golden standard, our study showed comparable sensitivity and specificity at a lower cut-off value [
33,
50]. Meanwhile, we conducted ROC analysis using the PHQ-9 as index of validity. The cut-off value 54 was also consistent with former research [
41]. The second is the study population since our research was conducted specifically on outpatients with type 2 diabetes in general hospitals across China. The psychosocial factors including the conception of illness, socioeconomic conditions, and awareness of mental symptoms may account for the difference in the cut-off score [
29].
The advantages of this study are as follows: First of all, this is the first study to examine the psychometric properties of WHO-5 in Chinese diabetic patients. Secondly, we used M.I.N.I. as the gold standard in determining the screening properties of WHO-5-C. M.I.N.I. is a structured diagnostic interview based on DSM-4 and ICD-10, which is more reliable for measuring depression than depression self-rating scales. Naturally, the study also has some limitations. First, the study was conducted in a third-class general hospital in China located in Beijing, and this may limit the generalization of the results across China, especially in primary care centers in rural or remote regions. Second, the participants were recruited between 2014 to 2015, the medication circumstances of which may be slightly different from the present. Thirdly, outcomes regarding the correlation between WHO-5-C and diabetes management self-efficacy and diabetes self-care behaviors are insufficient. The predictive validity of WHO-5-C should also be further investigated by reexamining the correlation with diabetic laboratory indicators, complications, mortality, and morbidity rate with a longitudinal study design.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.