Abstract
Background Stigma has been considered a significant barrier both in treatment, rehabilitation and help-seeking behaviours of people diagnosed with depression. This study aimed to validate the Depression Stigma Scale (DSS) to the Portuguese population, to assess the influence of the type of previous experience with depression on the levels of depression stigma, identifying depression stigma predictors in each type of previous experience with depression and to analyse the effects of stigma on help-seeking attitudes.
Methods A total of 1693 participants with a mean age of 47.2 (SD=18.17) completed the DSS, the Attitude Toward Seeking Professional Psychological Help, and a socio-demographic questionnaire. Participants were categorised into four comparison groups: no previous experience with depression (n=479), indirect experience with depression (n=661), direct experience with depression (n=137), and both direct and indirect experience with depression (n=416). We performed data analysis using SPSS 24.0.
Results The DSS demonstrated adequate psychometric properties. Levels of personal stigma were lower in people who had family and friends experiencing depression in comparison with individuals with no history of depression experience. Better attitudes towards help-seeking were evident in those with lower personal stigma and worse help-seeking attitudes were associated with higher perceived stigma in the indirect previous experience group.
Limitations There were some differences in the socio-demographic characteristics across groups and no data were collected on the duration of participant exposure to depression.
Conclusions Our study demonstrates the validity of DSS in the Portuguese population. The individual’s experience with depression influences the development of personal stigmatising beliefs and attitudes towards depression and plays a role in help-seeking behaviours. Addressing people’s experience of depression might be a practical way of reducing depression stigma and improve help-seeking behaviours.
Funding Funding for this study was provided by the European Community’s Seventh Framework Program (FP7/2007-2013) under grant agreement no. 223138. The funding partner had no role in the design of the study; in the data collection; in the analyses and interpretation of the data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.
1. Introduction
Globally, people diagnosed with depressive disorders comprise 4.4% of world population, and depression ranks as the most significant contributor to non-fatal health loss, accounting for 7.5% of all Years Lived with Disability (YLD) (World Health Organization, 2017). Moreover, although the underlying mechanisms are unclear, there is a well-documented association between depression and lower life expectancy (Cuijpers and Schoevers, 2004) that has been corroborated in a recent longitudinal population-based study which reported a higher rate of mortality over an 11-year follow-up period for individuals with psychological distress or major depressive disorder (Chiu et al., 2018).
Along with this high level of global burden, the treatment gap for people with depressive disorders worldwide is very significant, indicating that the majority of people do not access or receive the medical and psychological care they need (Thornicroft et al., 2017) despite the well-established return on investment associated with treatment (Chisholm et al., 2016).
One barrier to care access and adequate treatment might be the stigma associated with mental illness. There is evidence that such stigma, which is associated with feelings of shame, guilt, lack of self-esteem and social isolation (Rüsch et al., 2005; Lasalvia et al., 2013) as well as physical effects such as hypertension, elevated cortisol levels, and experiences of chronic pain (Sickel et al., 2019), negatively influences help-seeking and the therapeutic process (Rüsch et al., 2005; Busby Grant et al., 2015; Cornally and Mccarthy, 2011; Griffiths et al., 2008; Ibrahim et al., 2019; Kohls et al., 2017; Schnyder et al., 2017; Sickel et al., 2019). Thus, reducing stigma affecting people who have a mental illness such as depression is one of the main objectives of the Mental Health Action Plan 2013-2020 of the World Health Organization (World Health Organization, 2013).
Stigma is a major issue for people with a mental illness including those living with depression. For example, people with depression have reported community and institutional stigma similar to that experienced by people with psychosis (McNair et al., 2002) and a recent study found that 79% of the participants with major depressive disorders had experienced discrimination associated with their condition and between 20% and 37% of the participants have compromised their actions because of anticipated discrimination (Lasalvia et al., 2013).
Much of the recent research on depression and stigma has focused on two types of stigma: personal stigma and perceived stigma, personal stigma referring to one’s own beliefs about depression, and perceived stigma alluding to one’s beliefs about the attitudes of others (Griffiths et al., 2008). Recent studies have identified previous experience with depression as one of the leading predictors of personal stigma: a lower level of previous experience is associated with a higher level of personal stigma. In particular, studies have shown that participants with no previous experience with depression obtained greater personal stigma scores than those with previous experience with depression (Boerema et al., 2016). People with a history of depression showed less personal stigma than those with antecedents of parental depression and no previous experience with depression (Griffiths et al., 2008). In general, personal stigma decreases with depressive experience magnitude, with no significant impact on perceived stigma (Busby Grant et al., 2015).
Other identified predictors of personal depression stigma are lower education (Calear et al., 2011), and gender, with higher scores among male than females (Griffiths et al., 2008).
The association between prior experience with depression and perceived stigma is less clear, but some studies have reported some association between prior experience with depression and higher perceived stigma (Busby Grant et al., 2015; Griffiths et al., 2008).
Stigma plays an important role on help-seeking behaviours, and as it represents an essential vehicle for exploring and understanding patient delay and early action across a variety of health conditions including mental health care access (Cornally and Mccarthy, 2011), understanding stigma effects is crucial. Even though the importance of understanding the factors that influence help-seeking behaviours is well recognised, there remains a lack of research on the topic (Ibrahim et al., 2019).
A better understanding of stigma components and its predictors, has the potential of significantly improve stigma reduction interventions and promote help-seeking behaviours to decrease unmet need for treatment (Corrigan and Watson, 2002; Griffiths et al., 2008; Alonso et al., 2009; Busby Grant et al., 2015; Zimmermann and Papa, 2019).
The aims of this study were therefore threefold: to validate the Depression Stigma Scale to the Portuguese population, to assess the influence of the type of previous experience with depression on the levels of depression stigma in different socio-demographic groups, and to identify the relationship between depression stigma and help-seeking behaviour according to the type of previous experience with depression and its value as a predictor for help-seeking.
2. Methods
2.1 Participants and procedures
The sample collection for this study was part of the OSPI program – Optimising suicide prevention programs and their implementation in Europe, specifically within the application in Portugal (Hegerl et al., 2009; Kohls et al., 2017a), and the current study assumed a cross-sectional study design.
In Portugal, selection of participants was from the cable telephone network listed numbers, using the random digit dialling method to numbers belonging to Almada and Amadora municipalities. Trained interviewers conducted telephone contacts.
Almada and Amadora combined have above 350 000 inhabitants, being two of the most populous counties in Portugal. Due to its high population density, and the socio-economic diversity of its inhabitants, these municipalities were considered representative of the Portuguese population.
Of the 2009 participants in the OSPI research program (Hegerl et al., 2009, Coppens et al., 2013), 316 participants were excluded from the current study due to their exposure to the OSPI intervention which aimed to promote literacy on depression and, consequently, help-seeking behaviour. The data was collected between 2009 and 2010.
Previous experience with depression was self-reported by the participants upon two direct questions described in the supplementary file.
To investigate the relationship between past contact and stigma and help seeking, the remaining 1693 participants were categorised into four groups according to their level of past experience of depression. Group A (No experience) included 479 participants without experience of depression in themselves or a close contact. Group B (Indirect experience) comprised 661 participants who reported previous indirect experience of depression in a relative or close friend but not themselves. The 137 participants who had experienced depression themselves were allocated to group C (Direct experience). Group D (Direct and Indirect experience) included the 416 participants who had both experienced depression and had a relative or close friend with depression. Participant allocation by groups can be consulted in Figure 1.
2.2 Instruments
Participants completed a short socio-demographic questionnaire, the Portuguese translation of the Depression Stigma Scale (DSS) (Griffiths et al., 2004), and the Attitude Toward Seeking Professional Psychological Help: short form (ATSPH) (Fischer and Turner, 1970). The DSS comprises a 9-item Personal subscale and 9-item Perceived subscale with each item scored on a 5-point likert scale ranging from 0 (strongly disagree) to 4 (strongly agree). Higher scores indicate greater stigma. The ATSPH is a 10-item (range 0 to 30) with higher scores reflecting better help-seeking attitudes.
The socio-demographic questionnaire contained questions concerning the respondent’s sex, age, professional occupation and a question about previous depression diagnosis, a multiple-choice question with options to record the participant’s own experience and that of close relatives and friends (see supplementary file). As described above, the latter item was coded into the four categories: no experience of depression, indirect experience, direct experience, and both direct and indirect experience.
2.3 Data analysis
In addition to computing a continuous score for each DSS and ATSPPH item, we transformed the DSS score into percentages, in accordance to the original scale (Griffiths et al., 2004), with higher percentages indicating greater stigma levels. In the ATSPPH, the higher scores meaning positive help-seeking attitudes.
In order to study the psychometric properties of the DSS, we analysed floor-ceiling effects and response ranges, calculated Cronbach alphas, and conducted a principal component analysis.
We also calculated means and standard deviations for continuous variables and undertook ANOVAs followed by the Tukey HSD’s multiple comparisons test to compare stigma and help-seeking levels across groups. Student t-tests were applied to examine the differences between sexes within each contact group and Pearson correlations were used to examine the relationship between DSS scores and age.
Hierarchical linear regression analyses were performed to identify which variables were significant predictors of the depression stigma and help-seeking behaviour attitudes and the magnitude of the estimations. In the analysis of the predictors of depression stigma, gender and age was entered in the first model (Model 1), occupation in the second (Model 2), and depression experience in the final step (Model 3). In the analysis in which help-seeking was the criterion variable, depression experience was entered in the first step (Model 1), personal stigma in the second (Model 2) and perceived stigma in the final step (Model 4).
Sensitivity analysis was applied to the sample through stratified randomization according to occupation and analysis repeated.
IBM SPSS Statistics, Version 24.0 software package was used to conduct statistical analysis.
2.3 Ethical considerations
Trained interviewers explained in the first contact the objectives of the study and obtained verbal informed consent. Anonymity and confidentiality of the data collected were guaranteed.
The approval of the study protocol occurred in the scope of the OSPI program (Coppens et al., 2013; Kohls et al., 2017). In Portugal, the OSPI study was approved by the Medical Sciences Faculty of the New University of Lisbon’s ethics commission in May 2009.
3. Results
3.1 Sample description
As to gender and age, 53.6% were women and the mean age was 47.2 years (SD=18.17, range 18-90).
In Group A, participants had a mean age of 47.59 (SD=19.81), 42.2% were female; in Group B participants had a mean age of 44.31 (SD=17.52) and 51.7% were female; Group C presented a mean age of 55.49 (SD=17.56) and 64.9% female; finally, participants in Group D had a mean age of 49.15 (SD=16.29), and 66.8% were female (see table in supplementary file for more participant characteristics).
As the data used in this study was previously treated and used in other researches – as mentioned above – no missing data was detected in the dataset used.
3.2 Depression Stigma Scale psychometric characteristics
Principal component analysis with oblimin rotation showed two dimensions of the DSS, the first corresponding to the first nine items and the second including items 10-18 of the scale (corresponding to the Personal subscale and Perceived subscales respectively). We did not observe floor-ceiling effects. The percentage of participants endorsing each item ranged from 6-81% and 10-94% for the Personal and Perceived subscales respectively, with a mean stigma score of 39.6 (SD=14.03) for the Personal subscale and 56.0 (SD=15.40) for the Perceived subscale. Internal consistency was adequate for each of the scales: (Cronbach a=0.75 Personal subscale, α=0.69 Perceived subscale, α=0.77 Total scale). There was a Cronbach a=0.18 between Personal and Perceived depression stigma subscales. More information in the supplementary file.
3.3 Personal Stigma
As shown in Table 1, we observed significant differences in personal stigma scores between groups A and B (p<0.001), groups A and D (p<0.001), B and C (p<0.05), and C and D (p<0.001), with the lower means obtained in group D, followed by group B. We found a similar pattern of between-group effects for the female sub-sample.
Personal depression stigma as a function of previous experience and sex.
Gender differences were significant in group D, where males showed higher personal stigma means than females. There was a significant positive correlation between age and personal stigma in each of the previous experience of depression groups (group A: r=0.37, p <0.001; group B: r=0.30, p <0.001; group C: r=0.22, p <0.05; and group D: r=0.22, p <0.001), the higher the age, the higher the personal stigma. Personal stigma was higher among retired participants than any other occupation (M=46.43, SD=14.35; F=54.12, p<0.001). Pearson correlations are statistically different between group A and group C (z=1.69, p<0.05) and between group A and group D (z=2.45, p<0.01).
Hierarchical linear regression analysis (Table 2) indicated that gender and age were significant predictors of depression personal stigma after controlling for all other factors. Being employed, has a significant effect on personal depression stigma reduction. Controlling for other factors, participants with previous direct and indirect experience of depression (Group D) showed significantly lower personal stigma than participants from other groups, with the strongest effect for those with no experience of depression (Group A)
Regression models on the different types of experience, gender, age and occupation among personal depression stigma
3.4 Perceived Stigma
There were no significant differences in perceived stigma as a function of type of experience of depression or gender except perceived stigma was lower among male than female participants in Group D (Table 3).
Perceived depression stigma as a function of different types of previous experience and sex.
There was a statistically significant, albeit weak, negative correlation between age and perceived stigma for each of the previous experience of depression groups: group A r=-0.107, p <0.05; group B r=-0.130, p <0.05; group C r=-0.176, p <0.05; and group D r=-0.192, p <0.001.
There was also a main effect of occupational group on perceived stigma (F=12.96, p<0.001), with all pairwise groups comparisons reaching statistical significance except the contrast between employed and unemployed participants. Students had the highest level of perceived stigma (M=61.23, SD=12.03) and retired participants the lowest (M=53.41, SD=14.81).
A hierarchical regression analysis confirmed that previous experience with depression was not a significant independent predictor of perceived depression stigma (see Supplementary Table). After controlling for other variables, only gender and age were predictors of perceived stigma with perceived stigma being higher among females than males (β=1.42, t=4.16, p<0.05) and older age being associated with lower perceived stigma (β=-0.12, t=38.87, p<0.001).
3.5 Help-seeking attitudes
Personal stigma has a significant impact on help-seeking attitudes, with a r=-0.339, p <0.001, however no significance was found on the correlation between perceived stigma and help-seeking attitudes.
As shown in Table 4, analysis of variance followed by paired comparisons indicated there was a statistically significant difference in the help-seeking attitudes between group A and group B with the group with previous indirect experience having more favourable attitudes to help-seeking than the group without experience of depression. The females in groups A and D showed more favourable attitudes to help-seeking than their male counterparts, as shown in Table 4. Less positive help-seeking attitudes were weakly associated with older age in each group except group C (group A: r=-0.190, p <0.001; group B r=-0.08, p <0.05; group C r=-0.04, p = 0.65; and group D r=-0.105, p <0.05).
Help-seeking attitudes as a function of previous experience and sex.
Considering the model that included only the type of previous experience with depression in the hierarchical regression analysis (Table 5), groups B and D demonstrated better attitudes towards help-seeking than their counterparts with no previous experience of depression. However, after controlling for personal stigma (Model 2), the effect of previous experience was no longer significant, and higher levels of personal stigma predicted less positive help-seeking attitudes (see Table 5). Perceived stigma did not add significantly to the model (Model 4).
As shown in Table 6, personal stigma had a negative effect on help-seeking in all groups, proving to be a strong predictor independently of the type of previous experience with depression. On the other hand, perceived stigma was a significant predictor only for those with indirect experience of depression (group B) with higher perceived stigma being associated with less positive attitudes to help-seeking in this group.
Effects of age, gender, different types of previous experience, personal and perceived stigma on help-seeking behaviour.
Type of previous experience effects on the relationship between stigma and help-seeking attitudes.
3.6 Sensitivity analysis
Due to the occupational differences between Group C and other groups, we collected a more similar subsample in the other groups using the stratified randomization method with the occupational status as the stratification factor. This procedure resulted in a sample of 1053 individuals: 316 in group A, with a mean age of 54.1 (SD=19.07) and 42.1% females; 318 in group B, 54.0 (SD=18.95) years of age where 55.7% were female; and 288 in group D, 52.4 (17.48) years old and 66.7% female. A more detailed description of the subsample can be found in the supplementary file.
Personal depression stigma in this subsample in descending order were as follows: group A (M=45.54, SD=14.08), group C (M=41.95, SD=14.40), group B (M=40.26, SD=13.58), and group D (M=36.80, SD=14.74). There was a significant difference in personal stigma levels across groups (F(3,1049) = 19.72 p<0.001). The level of personal stigma in group D (direct and indirect experience) was lower than in each of the other groups (group A: p<.001; groups B, C: p<.05) and personal stigma in group B (indirect experience) was lower than in group A (no experience) (p<0.05).
In this subsample, there was were no significant differences in perceived stigma as a function of the type of previous experience (F(3,1049) = 1.28, p=0.28) with mean perceived stigma ranging from 53.82 (SD=16.00) in group C to 56.71 (SD=14.18) in group B. As in the total sample, the only difference observed was between genders in group D (t(287)= -2.66, p<0.05).
4. Discussion
4.1 Summary of main findings
The first aim of the study was to investigate the psychometric properties of the Portuguese version of the Depression Stigma Scale (DSS). The findings suggested that it is a suitable instrument for measuring both personal and perceived depression stigma in the Portuguese population.
The second study aim was to examine the influence of previous experience with depression on personal and perceived depression stigma. The absence of previous experience of depression (Group A) was associated with the highest level of personal stigma. Interestingly, having experienced depression oneself (Group C) was associated with higher personal depression stigma than with watching friends and family members suffer the disorder (Group B). People who had experienced depression both directly and indirectly (group D) showed lower levels of personal stigma than any other group, followed by those with close contact with a person with depression (Group B). In contrast, the type of previous experience with depression was not a predictor of perceived stigma, except for gender differences.
The third aim of the study was to examine the relationship between depression stigma and help-seeking behaviour according to the type of previous experience with depression and the value of personal and perceived stigma as a predictor of help-seeking. We found a negative effect of personal depression stigma on help-seeking attitudes for every type of previous experience with depression, whereas for perceived depression stigma effects were non-significant, except for females.
4.2 Discussion and literature
Although a number of studies have explored the effects of previous experience with mental disorders on stigma (Busby Grant et al., 2015), and on depression in particular (Griffiths et al., 2008; Wang and Lai, 2008), to our knowledge the current study is the first to systematically report on the gradient of exposure to depression.
4.2.1 DSS validation
The Portuguese version of the Depression Stigma Scale (DSS) showed adequate psychometric properties, suggesting that it was an appropriate instrument for the current investigation of the associations between stigma, help-seeking and experience and a promising scale for use in future studies in the Portuguese population. Although, internal consistency in the perceived subscale was lower than that obtained in its original form (α=0.88) (Griffiths et al., 2004), and in the Dutch version (α=0.82) (Boerema et al., 2016), it was nevertheless satisfactory. Internal consistency for the personal subscale, was similar to that reported in previous studies. However, future research is needed to explore the convergent validity of the scale.
Consistent with previous studies (Calear et al., 2011; Griffiths et al., 2008), perceived stigma scores were higher than personal stigma. This may represent a real effect or alternatively it might raise the possibility that perceived stigma is actually a proxy for personal stigma without social desirability bias (Boerema et al., 2016): higher perceived stigma corresponds to lower personal stigma.
4.2.2 Personal stigma
The finding of lower levels of personal stigma among those experiencing depression in other or self is consistent with previous studies that have reported a negative association between level of previous experience with depression and personal stigma (Griffiths et al., 2008; Busby Grant et al., 2016). However, the previous indirect experience group (B) had a significantly lower personal stigma mean than direct previous experience (C).
To our knowledge, no other research has compared the personal stigma between the direct and indirect previous experience with depression.
The gender differences observed in people in Group D who had experienced depression both directly and indirectly are in line with the findings of previous studies (Griffiths et al., 2008), but there were no significant observable differences in the other groups. In other words, in groups with a higher personal stigma, no gender differences were detected, suggesting that the experience of depression has a stronger effect on females.
Within the male group, individuals with no previous experience with depression (A) had higher personal stigma than those with previous indirect experience (B) and both previous direct and indirect experience (D). However, previous direct experience (C) had no significant effect on personal stigma reduction. We found similar results within the female group; yet, in this case, females with direct experience (C) showed the second-highest personal stigma means.
These results may provide an important insight into the effect of type of depression experience on stigma. It is possible that observing significant pain in someone close reduces personal depression stigma more than the more proximal experience of suffering depression oneself with individuals being more sympathetic and less judgmental about others than with themselves.
On the other hand, aging has a stronger effect on stigmatisation when people have experienced depression through a close other (group B) than when experiencing depression oneself (group C), raising the hypotheses of evanescent recall. Future research should take in account the time elapsed since the depression experience, because it would be interesting to understand the depression stigma reduction durability of the previous experience with depression.
It is legitimate to speculate that some internalisation of the discrimination experiences reported in previous studies (Lasalvia et al., 2013) may have occurred and therefore, working on stigma reduction among people with depression may be of particular clinical relevance.
4.2.4 Perceived stigma
On the other hand, the type of previous experience of depression was not a necessary predictor of perceived stigma, though men experiencing depression in self and others (group D) presented less perceived stigma then those who did not experience it (A) or experienced it only in others (B). There were also gender differences in both types of previous experience group (group D), with a higher mean perceived stigma among females. The explanation for these findings is unclear but suggests that when experiencing depression in self and others, men feel others are less stigmatizing, externalizing their own feelings.
4.2.5 Personal versus perceived stigma
In this study we found differences between personal and perceived stigma. In spite of a slightly positive correlation between these two subscales, whereas personal stigma increased with age, perceived stigma diminished with age. Also, women present a lower score in personal stigma show an higher score in perceived stigma.
We observed a positive correlation between age and personal depression stigma, with higher scores in older individuals, in agreement with previous results (Boerema et al., 2016) and the opposite effect on perceived stigma. This correlation was highest in individuals with no previous experience of depression for personal stigma when comparing with those with direct and both direct and indirect previous experience with depression.
Thus, an inverse correlation between these two constructs is suggested: the lower the personal stigma, the higher the perceived stigma, as previously reported in the literature (eg, Griffiths et al, 2008). It is also possible that perceived stigma scores reflect an oversensitivity to and a resulting overestimation of stigma in the community by those with lower levels of personal stigma.
4.2.6 Help-seeking
Personal depression stigma had a significant adverse effect on help-seeking attitudes in every group, in agreement with previous studies (Perlick et al., 2001). On the other hand, perceived stigma bear no effect on help-seeking attitudes in our study, contrary to what was previously observed in previous studies (Barney et al., 2006), except for those with indirect previous experience, where we observed a decrease on the likelihood of seeking professional help from those with higher perceived stigma.
4.2.7 Limitations
There are several limitations to this study.
Firstly, there are some differences in the sociodemographic characteristics across groups, namely group C. Secondly, phone interviews may have induced some socially desired answers. Thirdly, there is no information on the severity, duration and time of onset of the depressive episode, which could be new study variables for further research. Fourthly, the cross-sectional nature of the study, was a constraining factor limiting the precluding the exploration of other variables that could affect stigma levels, such as mental literacy promotion over time, the use of mental health care services, and experiences of discriminatory attitudes. Finally, depression severity at the time of assessment could influence the perceived stigma and it was not controlled. Future studies may benefit of its inclusion both to better understand the effects as well as to control it as possible cofounder. These limitations deserve future consideration in further research.
Data Availability
Data is available upon request to the corresponding author.
References
