Abstract
Background:
Depression is a major social concern in Japan. It is therefore necessary to develop a scale in Japanese that can assess depression literacy.
Aims:
The present study aimed to develop the Japanese version of the Depression Literacy Scale (D-Lit-J), and examined its validity and reliability.
Methods:
Three groups were administered the D-Lit-J, including 117 first-year university English literature students, 112 first-year medical school students, and 53 psychiatrists. Among these, 112 (95.7%), 112 (100%), and 29 subjects (54.7%) returned completed questionnaires, respectively. The total D-Lit-J scores were compared between the three groups to assess known-group validity, and internal reliability was examined by calculating Cronbach’s alpha coefficients. Medical students were asked to complete the questionnaire a second time, 3 weeks later (11 students did not respond), to assess the test–retest reliability using the intra-class correlation coefficient.
Results:
The total D-Lit-J scores (mean ± SD) were 7.61 ± 4.18, 9.51 ± 4.37, and 17.7 ± 3.15, for English literature students, medical students, and psychiatrists, respectively, and there were significant differences between the three groups (p < .05). The Cronbach’s alpha coefficients ranged from .800 to .834 in all students, and was .764 in psychiatrists, revealing a good internal consistency. The intra-class correlation coefficient of the scale was .769.
Conclusions:
The D-Lit-J showed a credible known-group validity, with good internal and test–retest reliabilities. Additional studies with a greater variety of subjects and that examine concurrent or discriminant validity will be necessary in the future.
Introduction
Depression is one of the most common mental disorders. Indeed, depressive disorders have been reported to be a leading cause of increases in the number of all-age years lived with disability, which is indicative of non-fatal health loss (GBD 2017 Disease and Injury Incidence and Prevalence Collaborators, 2018). An estimated 4.4% of the global population was estimated to have depression in 2015. Depression can be long-lasting and recurrent, and substantially impairs an individual’s ability to function at work or at school, and to cope with daily life (World Health Organization, 2017).
The lifetime prevalence of major depressive disorder (as defined by the Diagnostic and Statistical Manual of Mental Disorders, fourth edition) among the Japanese population is estimated to be 5.7% (Ishikawa et al., 2018). In 2017, a nationwide survey in Japan showed that the number of patients with depression has increased 2.9-fold in 21 years (Ministry of Health, Labour and Welfare, 2017). Accordingly, the Ministry of Health, Labour and Welfare in Japan has recognized depression as a critical health issue that requires early detection and rapid treatment. For this, it is important for both health care professionals and the general public in Japan to acquire a good depression literacy.
Depression literacy is defined as the ability to recognize depression and make informed decisions about treatment (Deen & Bridges, 2011; Wang et al., 2007). Among the four existing scales to evaluate depression literacy (Wei et al., 2016), the Depression Literacy Scale (D-Lit; Griffiths et al., 2004) has been used in a wide range of subjects (Darraj et al., 2016; Fonseca et al., 2017; Gulliver et al., 2012; Ibrahim et al., 2019). In addition to the original English version, the D-Lit has been translated into Arabic (Darraj et al., 2016), Bengali (Arafat et al., 2017), Greek, Italian (Kiropoulos et al., 2011), Korean (Bernstein et al., 2020), and Persian (Bahrami et al., 2019). Its Cronbach’s alpha coefficients have been reported to be .7 in athletes, with a test–retest reliability of .71 (Gulliver et al., 2012).
In Japan, several researchers have used their own scales to assess depression literacy, for example, investigations into the effects of stress and depression literacy interventions on improving symptoms and knowledge of depression (Imamura et al., 2016) and assessing public beliefs in and stigma toward people with depression (Yokoya et al., 2018). Depression literacy has also been examined using a measure developed by Jorm et al. (1997); this measure has been translated into Japanese (Nakane et al., 2005), revised by the authors, and used by Japanese researchers to assess depression literacy (Kaneko & Motohashi, 2007; Kubo et al., 2018; Nakamura-Taira et al., 2018; Ojio et al., 2015, 2019). However, these studies did not report the reliability or validity of this measure. It is therefore necessary to develop a valid and reliable scale in Japanese to assess depression literacy in a variety of subjects. To this aim, we developed a Japanese translation of the D-Lit (D-Lit-J) and examined its validity and reliability. Considering that medical undergraduates are more likely to obtain depression literacy than other students (Amarasuriya et al., 2015) and that the understanding of depression differs according to health training background (Hickie Am et al., 2007), we compared D-Lit-J scores between English literature students, medical students, and psychiatrists.
Methods
Development of the D-Lit-J
The 22-item D-Lit assesses depression literacy. Participants are asked to respond to each item with ‘True’, ‘False’, or ‘Don’t know’. Correct responses receive a score of 1, and incorrect or ‘Don’t know’ responses receive a score of 0. The total score ranges from 0 to 22, and higher scores indicate a higher depression literacy.
We developed the first D-Lit-J as follows: After receiving permission from the author of the original D-Lit (Griffiths et al., 2004), one of the authors (T.I.), a Japanese nursing researcher, and two English language lectures in the university translated the D-Lit into Japanese. Next, 12 researchers who specialize in public health and nursing reviewed the translations, compared them to the original English version, and made further suggestions for content and language-associated modifications. Then, the credibility of the translation was assured by back-translation by a bilingual Japanese psychiatrist. The original lead author of the D-Lit (Griffiths, K.M.) then checked the expressions used in the back-translated version that were considered difficult for Japanese people to understand, and, after discussions with her, we added supplementary explanations to the words and phrases. To assess the face validity of the translated measure, several health care professionals, non-professionals, and psychiatrists who did not know the purpose of the present study reviewed the draft, and amended it to obtain the final version of the D-Lit-J.
For item 17, ‘Counseling is as effective as cognitive behavioral therapy for depression’, the term ‘major depressive disorder’ (‘Dai-Utsu-byo’ in Japanese) was used instead of ‘depression’ (‘Utsu’ in Japanese), which was used in the original English version. This is because the Japanese word ‘Utsu’ encompasses various mental statuses relating to depressive mood (Ohmae, 2012), and we wished to avoid ambiguity.
Participants
Figure 1 shows a flow diagram of the study procedure. Between December 2018 and April 2019, two groups of first-year university students majoring in English literature in Tokyo (n = 117) and in medicine in Chiba (n = 112), as well as 53 psychiatrists at 4 mental health clinics affiliated with university hospitals in Tokyo, Chiba, and Saitama, were asked to participate in the study. The university students were asked to complete the questionnaire during their classes, which consisted of the D-Lit-J and questions regarding their demographics. Complete questionnaires were received from 112 students majoring in English literature (95.7%) and 112 medical students (100%). The medical students were asked to complete the questionnaire a second time after 3 weeks to assess the test–retest reliability; 101 of them (90.2%) responded. The same questionnaire was mailed to 53 psychiatrists, of whom 29 returned it (response rate of 54.7%). Before conducting the study, the study plan was reviewed and approved by the Ethics Committee of Juntendo University Faculty of Medicine, Tokyo, Japan (approval number 20181128, November 28, 2018; and approval number 2018192, March 19, 2019).

The study procedure.
Statistical analysis
The total D-Lit-J scores were compared among the three groups (English literature students, medical students, and psychiatrists) to assess the known-group validity using analysis of variance. The internal reliability was examined by calculating Cronbach’s alpha coefficients within the three groups and for all students combined. The intra-class correlation coefficient (ICC) was calculated to assess the test–retest reliability within the 101 medical students who returned a second completed questionnaire. Data analysis was conducted using Statistical Package for Social Science version 25 (IBM, Japan Inc., Tokyo, Japan).
Results
Table 1 shows the demographic characteristics of the English literature students, medical students, and psychiatrists.
Characteristics of the three groups of respondents.
Number (%).
Mean (SD).
The total D-Lit-J scores (mean ± SD) are shown in Table 2, along with the Cronbach’s alpha coefficient for each group. The scores were significantly different between the three groups, increasing in the order of English literature students, medical students, and psychiatrists (Figure 2), which indicated that the D-Lit-J had a good known-group validity. The Cronbach’s alpha coefficients of the D-Lit-J were .800 to .842 in the combined student groups and .764 in the psychiatrist group, which indicated a good internal reliability of the scale (Table 2). The scores of the two questionnaires completed by the 101 medical students were significantly correlated, with an ICC of .769, which indicated a good test-retest reliability (Figure 3).
D-Lit-J scores in English literature students, medical students, and psychiatrists.
Note. SD = standard deviation.

Comparison of D-Lit-J scores among the three groups.

Relationship of the D-Lit-J total scores between the first and second sessions in the 101 medical students.
Table 3 shows the actual wording of the 22 items of D-Lit-J, along with the correct response rates in each group. The percentage of correct answers ranged from 4.5% to 75.9% in English literature students, 0.9% to 84.8% in medical students, and 20.7% to 100% in psychiatrists. Among all students, an especially low correct response rate was seen for item 17 (‘Counseling is as effective as cognitive behavioral therapy for depression’) and item 21 (‘Antidepressants are addictive’). In psychiatrists, the correct response rate for item 17 was 20.7%, which was the lowest correct response rate of all 22 items within this group.
Number and percentage of participants who correctly answered D-Lit-J items.
Discussion
The D-Lit is a widely used instrument that assesses depression literacy. We translated the D-Lit into Japanese and investigated its validity and reliability in university students and psychiatrists. The known-group validity test showed that the D-Lit-J was able to recognize those who are expected to have a greater knowledge of depression, i.e., psychiatrists. The scores were significantly different between English literature students and medical students. Similarly, the depression literacy of medical students was higher than that of those in other disciplines (Amarasuriya et al., 2015).
The Cronbach’s alpha coefficients demonstrated a good internal consistency in the present study (.764–.842), which supports previous findings from Bahrami et al. (2019), Darraj et al. (2016), and Gulliver et al. (2012). Thus, the D-Lit-J, as with versions in other languages, seems to have an acceptable internal consistency. In addition, the ICC was .769 in the present study, which indicates a good test–retest reliability (Koo & Li, 2016).
The correct response rates to item 17 (‘Counseling is as effective as cognitive behavioral therapy for depression’) and item 21 (‘Antidepressants are addictive’) were below 10% among all students. Low correct response rates to these two items have also been observed in previous studies. For example, Fonseca et al. (2017) reported correct response rates of 10.3% and 10.8% for items 17 and 21, respectively, in women during the perinatal period in Portugal. The correct response rate for item 17 was 13.8% in Iranian female high school students (Bahrami et al., 2019). These two items contain terminology that is specific to the field of psychiatry, which could mean that they are more difficult for non-experts to understand and, thus, correctly respond to.
The psychiatrists also had a low correct response rate for item 17, despite our choice to use the more specific term ‘major depressive disorder’ (Dai-utsu-byo) in the place of ‘depression’ (Utsu). This result could indicate that psychiatrists in Japan consider counseling to be as effective as cognitive behavioral therapy for depression; indeed, authors such as Barth et al. (2013) have reported that both these measures can effectively reduce depressive symptoms. Despite the low rate of the correctness of item 17 in psychiatrists in Japan, this item is retained in D-Lit-J, because D-Lit including item 17 is used in various samples around the world, and the international comparison becomes possible by using D-Lit-J including item 17 as a universal item. When interpreting the results of D-Lit-J, however, it is necessary to consider the existence of this item.
The present study demonstrated that the D-Lit-J is a reliable and valid tool to assess knowledge of depression; it could therefore serve as an instrument to evaluate the efficacy of intervention programs to promote depression literacy. This study has several limitations, however. First, participants were limited to university students and psychiatrists from the Tokyo metropolitan area of Japan. Therefore, the generalizability of our findings to the rest of Japan is unclear. Further studies will be needed to conduct the study in the general population with different educational backgrounds throughout the Japanese population. Second, English literature students in this study were all females, whereas medical students and psychiatrists consisted of both males and females. Since the prevalence of depression differs according to gender and age (Ghadirian & Sayarifard, 2019; Kheirabadi et al., 2021), it should be investigated whether there is any variation of depression literacy with gender and age of the respondents in a further study. Third, the validity was examined only for the known group; therefore, other types of validity, such as concurrent validity and discriminant validity, should also be studied.
Conclusions
The D-Lit-J showed a credible known-group validity with good internal and test–retest reliabilities. Additional studies with participants that have different educational backgrounds throughout the Japanese population and examinations of concurrent or discriminant validity will be necessary.
Footnotes
Acknowledgements
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
