Stigma, coping strategies, distress and wellbeing in individuals with cervical dystonia: a cross-sectional study

ABSTRACT Cervical dystonia (CD) is a movement disorder which causes sustained muscle contractions in the neck leading to abnormal postures and repetitive movements. As it is a highly visible condition, people with CD can experience stigma, which may lead to unhelpful coping strategies and increased psychological distress. This study investigated whether adaptive and maladaptive coping strategies mediate the relationship between stigma and psychological outcomes in people with CD. A total of 114 adults with CD completed measures of stigma, coping, health-related quality of life (HRQOL), psychological distress (depression, anxiety, stress), and psychological wellbeing at one time point. Participants’ levels of distress were high, compared to the general population. Correlational analyses showed increased stigma and maladaptive coping (e.g. substance use, behavioural disengagement) were both significantly related to increased distress, lower wellbeing and lower HRQOL, whereas higher adaptive coping (e.g. acceptance, humour) was only related to higher wellbeing. In a parallel mediation model, maladaptive coping strategies mediated the relationship between stigma and distress, HRQOL and wellbeing, but adaptive coping strategies did not. These findings suggest that maladaptive coping may play an important role in explaining the relationship between stigma and some aspects of distress and wellbeing in CD. Interventions which focus on reducing different aspects of maladaptive coping may be helpful to improve wellbeing as well as reducing stigma.


Introduction
Dystonia is the third most prevalent movement disorder and causes involves involuntary, sustained muscle contractions leading to repetitive movements and abnormal postures (Albanese et al., 2018;Defazio, 2010).In addition to physical impacts, psychological difficulties are common with lifetime prevalence estimates of anxiety and depression between 12%-71% (Kuyper et al., 2011).Psychological distress may be both a result of the CONTACT Fiona J R Eccles f.eccles@lancaster.ac.uk,Division of Health Research, Lancaster University, Sir John Fisher Drive, Lancaster, LA1 4AT, UK underlying neurobiology associated with the condition, and secondary to the impact of living with a chronic health condition (Conte et al., 2016).
A range of factors have been shown to predict distress and health-related quality of life (HRQOL), including body image (Gündel et al., 2001;Jahanshahi, 1991;Lewis et al., 2008;Page et al., 2007;Pekmezovic et al., 2009), social participation (Ben-Shlomo et al., 2002;Jahanshahi, 1991), self-esteem (Ben-Shlomo et al., 2002;Jahanshahi, 1991;Lewis et al., 2008) and coping strategies (Gündel et al., 2001;Ioannou & Altenmüller, 2014;Jahanshahi, 1991).In studies of other movement disorders, such as Parkinson's and multiple sclerosis (MS), stigma has also been found to play a key role in HRQOL, distress and wellbeing (Ma et al., 2016;Valvano et al., 2016;Verity et al., 2020).Stigma is a vital aspect of health research and was originally defined as a feeling of being shunned by others for possessing a distinguishing characteristic that is less valued than the social norm (Goffman, 1963).Visible characteristics, such as tremors or muscle contractions in dystonia, can result in stigmatising social experiences (Morgan et al., 2019;Papathanasiou et al., 2001) and studies have found the public hold stigmatising attitudes towards people with dystonia (Rinnerthaler et al., 2006).However, the role of stigma in relation to psychological outcomes for people with dystonia is poorly understood with both significant (Ben-Shlomo et al., 2002) and non-significant (Basurović et al., 2012) findings previously reported, but both studies used a stigma scale not validated for dystonia.
The 'minority stress paradigm' (I.H. Meyer, 2003), conceptualises stigma as a stressor above and beyond what non-stigmatised people endure, leading to difficulties with mental health.This paradigm was initially developed to capture the experience of lesbian, gay and bisexual individuals and to explain how a stressful social environment leads to psychological distress.More recently it has been applied to other minoritised groups, such as autistic people (Botha & Frost, 2020) and those with disabilities, including those with chronic illness (Lund, 2021).Given the experiences of stigma of people with dystonia (Morgan et al., 2019;Papathanasiou et al., 2001), and the negative attitudes held by others (Rinnerthaler et al., 2006), the model seems appropriate to investigate here.Hatzenbuehler (2009) further developed the minority stress ideas to propose a psychological mediation framework which explains how stigma leads to psychological processes, which in turn creates psychological distress.Hatzenbuehler (2009) proposed three main processes as mediators: social and interpersonal factors such as social isolation and reduced social support; cognitive factors including hopelessness, pessimism and negative self-schemas; and emotional regulation and coping, the latter being the focus of the present study.
Coping can be defined as the behavioural, emotional and cognitive strategies employed to manage certain stressors (Folkman et al., 1986) and has an important role to play in the adaptation needed due to changes in physical health (Brands et al., 2012).As Hatzenbuehler (2009) proposed, empirical work has suggested experiences of stigma can influence the coping strategies adopted for those living with physical health conditions.For example, avoidance of social situations is a common coping response to stigma in visible physical illnesses (e.g.(Maffoni et al., 2017;Mayor et al., 2022;H. Smith et al., 2015).People may also try to conceal the physical illness where this is possible (Maffoni et al., 2017;Mayor et al., 2022;H. Smith et al., 2015).However, adaptive coping strategies can include selfacceptance and self-advocacy as well as selective disclosure and accessing support from appropriate others (Mayor et al., 2022;H. Smith et al., 2015).When considering specific coping strategies to manage physical illness, higher stigma has been found related to higher resignation/giving up coping (a passive response), lower confrontation coping (taking active steps to manage the illness) (Luo et al., 2023;Tang et al., 2023) and higher avoidant coping (Earnshaw et al., 2018;Luo et al., 2023).Resignation coping mediated the relationship between stigma and lower quality of life for those with a diabetic foot ulcer (Luo et al., 2023) and avoidant coping mediated the relationship between stigma and health outcomes for those with HIV (Earnshaw et al., 2018).Stigma associated with depression has also been associated with increased substance use (Wang et al., 2018).
Coping strategies have been classified in various ways within the literature, for example 'emotion focussed', 'problem focussed', 'appraisal focused', 'approach', 'avoidance', 'primary' and 'secondary', but with no consensus on the most appropriate classification (Skinner et al., 2003).A recent review of the factor structures found in a commonly used coping measure, the brief COPE (Carver, 1997), found a 2-factor model was the most common, although the precise factors differed across studies (Solberg et al., 2022).While different coping strategies can be more or less helpful depending on the situation, nonetheless there are some coping strategies that tend to result in more negative outcomes (strategies such as helplessness or rumination) and others that may often lead to more positive outcomes (strategies such as planning or meaning making) (Skinner et al., 2003).Some studies undertaking factor analyses of coping questionnaires have indeed found these 'adaptive' and 'maladaptive' factors (e.g.Mugavero et al., 2009;Webb Hooper et al., 2013), while other studies have presumed this binary adaptive-maladaptive structure from a theoretical perspective (e.g.B. Meyer, 2001), including when investigating the relationships with stigma and distress (Tran & Lumley, 2019).In particular, maladaptive coping strategies have a greater relationship with distress, whereas adaptive strategies have a stronger relationship with wellbeing (B.Meyer, 2001).In dystonia, maladaptive coping has previously been shown to be related to higher levels of depression and anxiety (Gündel et al., 2001;Ioannou & Altenmüller, 2014;Jahanshahi, 1991;Scheidt et al., 1995), although the influence of coping on wellbeing in dystonia has not been studied.Wellbeing is usually conceptualised as a combination of functioning well and positive affect (Deci & Ryan, 2006), and is an important predictor of physical health in longitudinal research (Boehm et al., 2011;Xu & Roberts, 2010), so is included in the present study.
In summary, previous studies have found relationships between maladaptive coping and psychological distress in dystonia, with the relationship between stigma and distress being unclear.However the relationship between stigma, coping and distress has not been studied in dystonia, nor has the relationship with wellbeing.Consequently, this study investigates the relationship between stigma, coping and psychological outcomes (psychological distress, wellbeing and HRQOL) and also investigates whether coping mediates the relationship between stigma and psychological outcomes in a cohort of people with cervical dystonia (CD), the most common type of dystonia.

Design
This study was a cross-sectional survey using quantitative measures.Bivariate correlations were carried out to test the predicted relationships, followed by a parallel mediation analysis to test the theoretical model that coping strategies mediate the relationship between stigma and psychological distress (anxiety, depression, stress), wellbeing and HRQOL.

Participants
Participants were recruited via two dystonia charities: Dystonia UK (https://www.dystonia.org.uk/) and Dystonia Ireland (http://www.dystonia.ie/).Recruitment took place between 3rd September, 2019 and 27th January, 2020 via advertisements on the charities' social media profiles.Two participants completed paper copies and the rest completed it online.
A priori effect size calculations suggested for a medium effect (0.36) in both paths (stigma to coping (α) and coping to psychological outcomes (β)) using a bias-corrected bootstrap mediation model, approximately 71 participants were required for 80% power (Fritz & Mackinnon, 2007).A total of 114 complete datasets were available for analysis.
Individuals were eligible to take part if they self-reported a diagnosis of cervical dystonia; were aged 18 or over; could complete the measures alone or with support; had sufficient understanding of written English to take part.Potential participants were excluded if they had dystonia following a serious injury and/or had another significant illness/condition that affected their visible appearance.

Procedure
Ethical approval was obtained for this study via the Lancaster University Faculty of Health and Medicine Research Ethics Committee on 30 th August 2019.The study was designed in consultation with two members of Dystonia UK who provided feedback via email on the content and accessibility of the survey.The study was advertised via Dystonia UK and Dystonia Ireland and participants completed a (mainly online) survey using Qualtrics, a web-based survey tool, including giving written consent via this tool.Paper copies were provided on request.

Materials
A demographics questionnaire asked about age, gender, ethnicity, work status, relationship status and living arrangements.A clinical information questionnaire asked about age of onset, duration, time since diagnosis, medication, disease severity, whether the individual was receiving treatments and had any other health conditions.
The predictor variable, stigma, was measured by the 24-item Stigma Scale for Chronic Illness (SSCI) which was developed for people with chronic neurological disorders (Rao et al., 2009).The total score measures overall stigma (range 24-120, α = 0.96).
The two mediator variables, adaptive coping and maladaptive coping, were both measured using the brief COPE (Carver, 1997), adopting the two-category model (B.Meyer, 2001), which has subscales for adaptive (8 items) and maladaptive (6 items) coping strategies.Adaptive coping scores range from 16-64 and maladaptive from 12-48.
The second outcome measure, the Warwick-Edinburgh Mental Well-being Scale (WEMWBS) (Tennant et al., 2007), is a 14-item validated measure of wellbeing with a total score range of 14-70, α = 0.89.
The measure of HRQOL was the Cervical Dystonia Impact Profile (CDIP-58) (Cano et al., 2004) which is a 58-item scale measuring the impact of CD across eight physical and psychosocial domains, α = 0.92.

Data analysis
One hundred and forty-three participants opened the survey and completed the consent process, but 29 responses were removed due to large amounts of missing data (e.g.only demographics present).Of the remaining cases (n = 114), none were missing more than two data points from any measure, so they were retained.Little's MCAR test (Little, 1988) was not significant on these remaining cases, so there was no evidence from this analysis to reject the null hypothesis that data were missing completely at random (MCAR), although this analysis can have low power (Baraldi & Enders, 2010) and a non-significant result alone does not prove the data are missing completely at random (Rhoads, 2012).However, visual inspection indicated the missing values were spread across the questionnaires, apart from one question on the CDIP-58 asking how often the participant had felt 'down', which was omitted by nine participants.The mean HADS depression score for these participants was lower than for those who had completed it (5.50 versus 7.19, p >.05) suggesting those with lower mood might be less likely to answer this question; the difference was not statistically significant.Given the very small amount of missing data for the 114 retained participants (<0.5% of all the datapoints) mean substitution was used to impute missing values to enable maximum use of the data collected.No differences were found in demographics and clinical characteristics for those who completed the full set of measures and those who did not.Sensitivity analyses were completed by running the analyses with only those participants with complete raw data (n = 64) and comparing them with the full imputed dataset (n = 114) and no differences were found in any of the significant and non-significant results so the analyses on the full 114 participants are reported here.
Data distributions were inspected visually for normality using histograms.The 'maladaptive coping', 'depression', 'anxiety' and 'stress' data were all skewed towards lower scores, so Spearman's ρ correlation coefficients were calculated to assess the direction and strength of relationships between predictor, mediator and outcome variables.Before conducting the mediation analyses, the linear and multiple regression scatterplots corresponding to each analysis in the mediation were also visually inspected to test the assumptions of linearity and homoscedasticity of residuals, with standardised residuals plotted against standardised predicted values (Field, 2018).Q-Q plots were used to assess the assumption of normality of error distributions (Field, 2018).All relationships appeared to respect the assumptions of linearity, homoscedasticity of residuals and normality of error distributions Finally, a series of mediation analyses were conducted using Hayes' Process Tool (Hayes, 2018) with 5000 bootstrap samples.In all models, stigma was the independent variable and two types of coping strategies (adaptive and maladaptive) were tested together as parallel mediators.The outcomes were HRQOL, wellbeing, stress, anxiety and depression.The models were run with 95% confidence intervals for the indirect effects (in line with the power calculation described above).However, given no correction in significance has been made for multiple testing, the 99% confidence intervals were also checked to indicate the robustness of the findings.

Results
A summary of participant characteristics is included in Table 1.The sample was mainly female (82%) and White (97%), with mean age of 52.39 years.Half the

Correlational analyses
Non-parametric bivariate correlations indicated that the relationships between all the scales, apart from the adaptive subscale of the Brief COPE, were significant at the p < 0.001 level with medium to large effect sizes (see Table 3).Both stigma and maladaptive coping correlated with the outcome measures as expected (higher stigma/maladaptive coping with higher psychological distress and lower HRQOL and wellbeing) and they correlated with each other.However, higher adaptive coping only correlated with higher wellbeing (p <.001).

Mediation regression analyses
For a summary of results including all path coefficients and effect sizes with 95% confidence intervals see Table 4 and Figure 1.
In all models using 95% confidence intervals maladaptive coping mediated the relationship between stigma and the outcomes (anxiety, depression, stress, HRQOL and wellbeing) but adaptive coping did not.In all models the direct effect between stigma and the outcome remained significant when including the mediational effects.These significant and non-significant results were the same when using 99% confidence intervals.X=predictor, M=mediator and Y=outcome; c' = direct effect of X on Y, controlling for M; c = total effect of X on Y; ab = mediated effect; CI = confidence interval; CSIE: completely standardised indirect effect (a measure of effect size).

Discussion
The current study aimed to examine associations between stigma, coping and psychological outcomes for people living with CD.The rates of moderate-severe, depression, anxiety and stress in this study (48.3%, 55.3% and 36.8%respectively) were perhaps somewhat high compared with some previous CD studies with rates of depression ranging from 16.4%-52.3%(Gündel et al., 2001;Jahanshahi, 1991;Jahanshahi & Marsden, 1990;Mahajan et al., 2018;Smit et al., 2016) and anxiety ranging from 21.7%-50% (Gündel et al., 2001;Mahajan et al., 2018;Smit et al., 2016).The high rates may be partly due to the measures used, as the DASS refers to population norms rather than clinical cut-offs although lifetime prevalence rates of distress for people with dystonia are known to be high (Zurowski et al., 2013) and the levels of wellbeing were low when compared to a previous dystonia sample (Sandhu et al., 2016).Stigma (mean = 61.33)was also high compared to other populations such as stroke (mean = 45.21) and multiple sclerosis (mean range 36.45-47.4)(Anagnostouli et al., 2016;Deng et al., 2019).However, despite these high means, participants' distress was scored across the range, from low to high (see Table 2), and therefore we argue the results here are not only applicable to those with high levels of distress.
The results supported the hypothesis that stigma is associated with maladaptive coping.Feeling devalued by society through either direct experience of stigma, or the internalising of others' attitudes, can lead to an increased use of maladaptive coping strategies with the aim of alleviating levels of distress (Hatzenbuehler, 2009;Tran & Lumley, 2019).However higher stigma was not found to be associated with lower adaptive coping, which contrasts with one study with people with mental health difficulties (Chronister et al., 2013) but concords with another in HIV (Rinehart et al., 2019).Additionally research looking at the mental health of students found that internalised stigma was also not correlated with adaptive coping (Tran & Lumley, 2019).It may be the case that, although stigma may increase the likelihood of adopting maladaptive coping strategies, this does not mean that it will decrease the use of any existing adaptive coping strategies.
As hypothesised, and in line with previous research in dystonia, stigma was associated with lower HRQOL (Ben-Shlomo et al., 2002).Stigma was also associated with increased psychological distress and lower wellbeing as seen in other conditions (Anagnostouli et al., 2016;Ma et al., 2016;Simpson et al., 2014;Verity et al., 2020).Similarly, maladaptive coping was found to be associated with increased levels of distress as previous dystonia research has also shown (Gündel et al., 2001;Ioannou & Altenmüller, 2014;Jahanshahi, 1991;Scheidt et al., 1996).Novel findings were identified of significant associations between maladaptive coping and lower HRQOL and lower wellbeing.
In contrast, adaptive coping was not found to be significantly associated with measures of psychological distress (stress, anxiety or depression) or HRQOL.It was however, associated with wellbeing.These findings fit with a meta-analysis which found that the association between adaptive coping and distress was weaker than the association between maladaptive coping and distress (Aldao et al., 2010), suggesting that the different strategies are not equal in their contribution to, and maintenance of distress.Furthermore, research into the concept of psychological wellbeing has identified that this is more than an absence of psychological distress (Trudel-Fitzgerald et al., 2019).Despite there often being an inverse correlation between wellbeing and distress, these are often moderate at best (Baselmans et al., 2018;Ryff & Keyes, 1995;Ryff et al., 2006) and wellbeing and distress have distinct biological features (Rector et al., 2019).This might be why the findings for adaptive coping were specific to wellbeing and, whilst use of such strategies might not reduce distress, they may be an important consideration in improving overall wellbeing.
When examined through a psychological mediation framework, maladaptive coping was found to mediate the relationship between stigma and distress, wellbeing and HRQoL which is in line with previous research in other populations that have explored this model (Rinehart et al., 2019;Tran & Lumley, 2019).However, adaptive coping was not found to mediate the relationship between stigma and wellbeing in any of the psychological mediation models examined.This is contrary to previous research which found that adaptive coping mediated the relationship between stigma and psychological wellbeing or distress in students (Tran & Lumley, 2019), and that it played a role in reducing symptoms of depression (Herman-Stabl et al., 1995).Thus, the results from the current study only partially support the theory that coping strategies are processes which impact on wellbeing when trying to manage stress relating to stigma (Hatzenbuehler, 2009;Tran & Lumley, 2019).Rinehart et al. (2019) also found no evidence for the mediating role of adaptive coping between stigma and psychological distress in individuals with HIV.They suggested that adaptive coping may serve more as a moderator that dampens the adverse effects of stigma.Support for this idea was found in another study which showed that adaptive coping had a moderating effect between HIV-related stigma and medication adherence (Martinez et al., 2012).Additionally, positive reframing (one of the sub-categories of adaptive coping in the scale used in this study) has been shown to be a protective factor when trying to manage the stress associated with stigma in a study looking at a psychological intervention (Tshabalala & Visser, 2011).This indicates that there may be benefit in future research considering the separate coping strategies within the scale as opposed to grouping based on the two-factor model.Rather than the type of coping, how and when strategies are used may be more important when considering the impact on distress (Montel et al., 2009).

Clinical implications
As indicated here and elsewhere (Ben-Shlomo et al., 2002;Kuyper et al., 2011;Papathanasiou et al., 2001), people with dystonia can experience high levels of stigma and psychological distress.The results suggest that coping strategies may have an important role to play in the relationship between stigma and distress.Developing skills relating to coping have been shown to be beneficial to mental health (Bettis et al., 2017).These skills can be modelled and developed within psychological therapy both within group formats (S.K. Smith et al., 2015) and, individually using approaches such as Acceptance and Commitment Therapy (ACT).
On a more systemic level, initiatives that target the reduction of stigma relating to this condition may also decrease distress, improve wellbeing and enhance HRQOL.A review on interventions for reducing health related stigma suggested that multiple levels of involvement should be considered including within the community, with those around the individual and at government and structural levels (Heijnders & Van Der Meij, 2006).For people with CD this might include the involvement of experts by experience in the design of any policies, providing training to health professionals to increase awareness of this condition and increasing public awareness.

Limitations and further research
The majority of respondents were White British and female.While dystonia is more prevalent in females (Jankovic et al., 1991;Norris et al., 2016), this may suggest bias in the sample.Ethnicity is less frequently reported in studies on CD but in one study the estimated prevalence of white people with CD was 1.23 per 100,000 people and other races was 0.15 per 100,000 people (Marras et al., 2007).The low numbers of participants from ethnic groups other than White British/English/Welsh/Scottish/Irish means that the findings may not apply more universally.Advertisement was primarily online which may have biased the sample towards younger people, although the mean age (52.3 years) was similar to another UK study (56.1 years) (Lewis et al., 2008).Recruiting via clinics instead of community may help address these biases.
This study involved conducting several analyses without a formal correction to reduce the family-wise error rate and therefore concerns may be raised about a Type 1 error.However, all the key correlational findings were significant at p < .001,with medium to large effect sizes and therefore would have remained significant, even with a considerably adjusted alpha value.We also tested the mediation models using the strictest criterion possible in the SPSS Process tool (99% confidence intervals) and the findings from the 95% confidence interval analysis still remained.We therefore have confidence in these findings.Mean substitution was used to impute missing data which can lead to biased parameter estimates (Graham, 2009).However, the sensitivity analyses suggested that the findings were robust and most missing data were in the CDIP where one or two items were missing out of a total of 58, suggesting limited impact on the total score.Cross-sectional mediational studies have limitations in terms of inferred causality (Hayes, 2018) and it may be the case that relationships identified here are bi-directional and/or more complex.For example, studies investigating the predictors of stigma can include mood, anxiety and coping (Hanff et al., 2022), and there may be feedback loops, whereby lowered mood causes further perceptions of stigma, which further lower mood.Similarly, certain coping strategies may exacerbate stigma and hence contribute to further distress.Furthermore, as noted above certain forms of coping may act as a moderator of the relationship between stigma and distress (Martinez et al., 2012;Rinehart et al., 2019).A longitudinal methodology would be useful to explore any temporal and/or causal links between stigma and wellbeing in dystonia and further research should also investigate further the precursors to stigma and allow more complex interactions.Categorising the concept of coping into two factors also has its limitations as the complexities and nuances of coping strategies can remain overlooked (Skinner et al., 2003).Future larger studies may benefit from undertaking factor analysis to inform the structure of coping for people with dystonia and/or using different conceptualisations of coping.Coping with stigma specifically may be important to address, as the coping explored in the current paper focused more on coping with the illness.Exploring the experiences of coping in those with dystonia utilising a qualitative methodology could also be beneficial to inform the categorisation for future quantitative studies.

Conclusion
Stigma and maladaptive coping could be important predictors of distress, lower wellbeing and lower HRQoL for people with CD.Maladaptive coping partially mediates the relationship between stigma and distress and HRQoL.Interventions which focus on reducing different aspects of maladaptive coping and reduce stigma may be helpful to reduce distress.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Table 1 .
Sample demographics.(57) reported being in either full time or part time work, 19 (17%) were retired and 38 (33%) were not currently working.Mean age at diagnosis was 42.60 years, most participants (88%) had received botulinum toxin injections to manage the symptoms of cervical dystonia and 55% took other prescribed medication to manage symptoms.
Percentages are rounded to the nearest whole number, except for percentages less than one, which are rounded to the nearest 0.5%.participants

Table 2 .
Means, SDs and Cronbach's alpha of psychometric measures.