NOTICE: This is the author’s version of a work that was accepted for publication in Appetite. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Appetite, Vol. 61 (2013). doi: 10.1016/j.appet.2012.11.001 Gluten free diet adherence       1     Title: Gluten free diet adherence in coeliac disease: The role of psychological symptoms in bridging the intention-behaviour gap Short title: Gluten free diet adherence Authors: Kirby Sainsbury a Barbara Mullanb Louise Sharpe a Clinical Psychology Unit, the University of Sydney a School of Psychology, the University of Sydney b Corresponding author: Barbara Mullan Address: School of Psychology, the University of Sydney, NSW, 2006, AUSTRALIA Email [email protected] Phone: + 61 2 9351 6811 Fax: + 61 2 9036 5223 Gluten free diet adherence       2     Abstract This study examined the potential role of psychological symptoms in limiting the translation of positive intention into strict gluten free diet (GFD) adherence in coeliac disease (CD) within a theory of planned behaviour (TPB) framework. It was hypothesised that participants with more symptomatic psychological profiles would exhibit poorer adherence, primarily in the context of positive intentions. Coeliac disease participants (N = 390) completed online measures of gluten free diet adherence, psychological symptoms, coping behavior, and TPB items. Intention and behaviour were moderately correlated, confirming the existence of the intention-behaviour gap. Psychological symptoms accounted for additional variance over and above TPB variables in GFD adherence but not intention. Participants who failed to act on their positive intentions displayed more psychological symptoms and greater reliance on maladaptive coping strategies than those with consistent intention-behaviour relationships (p < .01). The heightened incidence of psychological symptoms in CD has a small but significant negative impact on the ability to translate positive intentions into strict adherence. Directions for future research including interventions to improve GFD adherence are discussed. Key words: coeliac disease, gluten free diet, theory of planned behaviour, intention-behaviour gap, psychological symptoms Abbreviations: CD = coeliac disease; GFD = gluten free diet; ANOVA = analysis of variance Gluten free diet adherence       3     Gluten free diet adherence in coeliac disease: The role of psychological symptoms in bridging the intention-behaviour gap Introduction Coeliac disease (CD) is a chronic autoimmune disorder involving intolerance for dietary gluten and affecting approximately 1% of the population (Green & Jabri, 2006; West et al., 2003). Gluten consumption in affected individuals causes atrophy of the small intestine and is associated with both gastrointestinal symptoms and the malabsorption of nutrients (Green & Cellier, 2007). Left untreated coeliac disease can lead to serious and potentially life-threatening long-term health complications including intestinal and bowel cancers, osteoporosis, and infertility (Green & Jabri, 2003; Rubio-Tapia & Murray, 2010). Importantly, with strict adherence to a gluten free diet (the only effective treatment for coeliac disease) reduction of symptoms and the risk of complications can be achieved (Anderson, 2008; Bai et al., 1997; Kemppainen et al., 1999; West, Logan, Smith, Hubbard, & Card, 2004). Thus, the ability to maintain strict adherence to a gluten free diet is of utmost importance in this population. A systematic review of 38 gluten free diet adherence studies (Hall, Rubin, & Charnock, 2009) found that strict adherence estimates ranged considerably (range = 36 – 90%; median = 70%), although comparisons were limited due to variability in the method of assessment and definitions of strict adherence, and the lack of replication of most study designs. The review found no consistent relationship between adherence and demographic or disease factors (Hall et al., 2009), and concluded that there was a need for more rigorous empirical and theory-based research into the demographic, cognitive, emotional and disease characteristics that influence adherence in coeliac disease, as well as calling for a more reliable means of assessing dietary adherence. Gluten free diet adherence       4     The theory of planned behaviour is a commonly used model in the psychological and behavioural health research which posits that the most proximal predictor of behaviour is intention to perform that behaviour (Ajzen, 1991). Intention is in turn influenced by three factors: attitudes – beliefs about the likely outcome of the behaviour; subjective norms – perceptions of the expectation of others to perform the behaviour; and perceived behavioural control – perceived level of control over performance of the behaviour, which also has a direct effect on behaviour (Ajzen, 1991). The theory of planned behaviour has been successfully applied to the prediction of a number of health intentions and behaviours including fruit and vegetable consumption (Kothe, Mullan, & Butow, 2012; Povey, Conner, Sparks, James, & Shepherd, 2000), breakfast consumption (Wong & Mullan, 2009), intention to consume foods with a low glycemic index (Goodwin & Mullan, 2009), and dietary and exercise behaviours in type 2 diabetes patients (White, Terry, Troup, Rempel, & Norman, 2010). In the only reported theory-based study in coeliac disease the theory of planned behaviour was also applied to the prediction of gluten free diet adherence, and accounted for 37% of the variance in intention to maintain a strict gluten free diet and 22% of actual adherence (Sainsbury & Mullan, 2011). These results were consistent with a meta-analytic review (Armitage & Conner, 2001) which found that the theory accounted for 39% and 27% of the variance in intention and behaviour. The theory of planned behaviour has been criticised for its failure to explicitly consider the non-rational components of decision-making (Conner & Sparks, 2005; Sharpe & Curran, 2006). That is, behaviour is assumed to be logical and goal-directed, while the unconscious, irrational determinants of behaviour including emotional states, psychological problems, and coping skills are considered only to the extent that they inform the development of beliefs (Ajzen, 2011). Further it has been suggested that while the theory reliably accounts for significant Gluten free diet adherence       5     variance in both intentions and behaviour, there is not a perfect relationship (r = 1) between them; a phenomenon known as the intention-behaviour gap (Sheeran, 2002). That is, a significant proportion of people do not act in line with their intentions: those who fail to act on their positive intentions (inclined abstainers) and those who act despite negative intentions to do so (disinclined actors). Research suggests that it is the first group of people who are primarily responsible for the intention-behaviour gap (Sheeran, 2002). In contrast, inclined actors and disinclined abstainers represent intention-behaviour consistency for people with positive and negative intentions respectively. Previous research has attempted to improve the predictive ability of the theory of planned behaviour and narrow this gap by considering post-intentional factors such as planning and self-regulation abilities (Mullan, Wong, Allom, & Pack, 2011; Norman & Conner, 2005; White et al., 2010). The heightened incidence of psychological symptoms in coeliac disease (Addolorato et al., 2008) is proposed here as a potential candidate for addressing both these criticisms. Numerous researchers have reported an increase in the incidence of psychological symptoms within coeliac disease (Addolorato et al., 2008), including depression (Addolorato et al., 2001; Addolorato et al., 1996; Ciacci, Iavarone, Mazzacacca, & De Rosa, 1998; Siniscalchi et al., 2005), anxiety (Addolorato et al., 2001; Fera, Cascio, Angelini, Martini, & Guidetti, 2003), and eating disorders (Karwautz et al., 2008), with a relationship between psychological symptoms and increased gastrointestinal symptoms also being evidenced (Hauser, Musial, Caspary, Stein, & Stallmach, 2007). Attempts to explicitly link psychological symptoms to adherence have, however, been limited. Specifically, higher depression and anxiety symptoms have been correlated with poorer gluten free diet adherence, although failed to account for unique variance in adherence (Edwards-George et al., 2009). The self-reported ability to follow a gluten Gluten free diet adherence       6     free diet despite changes in mood and stress, but not the presence of self-reported comorbid psychological disorders per se, was also associated with improved adherence (Leffler et al., 2008). Finally, a relationship between the presence of eating disorders and gluten free diet adherence, as assessed by self-reported transgressions and serological analysis, has been reported (Karwautz et al., 2008). No study to date has explicitly assessed coping behaviour in coeliac disease, although qualitative evidence suggests that coping may impact on adherence (Hallert, Sandlund, & Broqvist, 2003; Olsson, Hornell, Ivarsson, & Sydner, 2008). These preliminary findings, combined with the observation that patients suffering from depression are three times more likely to be non-compliant with medical treatment recommendations (DiMatteo, Lepper, & Croghan, 2000), suggest that assessing the impact of psychological symptoms and coping on gluten free diet adherence is warranted. A major limitation in the literature is that although there have been attempts at linking disease, demographic, and psychological factors to gluten free diet adherence, individual relationships have largely been assessed in isolation. Research elsewhere has demonstrated the negative impact of depression on compliance (DiMatteo et al., 2000), particularly in chronic illness populations (Safren, Gonzalez, & Soroudi, 2008). Maladaptive coping is also linked to increased psychopathology (Wingenfeld et al., 2009), with evidence from varying medical conditions that coping styles, specifically an active and internally focused coping style, are related to improved treatment adherence (Christensen & Johnson, 2002). Further limitations include the lack of a theoretical base for understanding gluten free diet adherence and the lack of a consistent and validated measure of gluten free diet adherence. This study was therefore designed to address these limitations by firstly applying a social- cognitive model (the theory of planned behaviour) to the prediction of gluten free diet adherence, Gluten free diet adherence       7     measured using a validated questionnaire. The second aim was to improve the predictive ability of the theory of planned behaviour and narrow the intention-behaviour gap by adding measures of psychological symptoms and coping behaviour. It was hypothesised that the theory components would account for significant variance in intentions and gluten free diet adherence, and that psychological symptoms and coping behaviour would add to the prediction of adherence over and above intention and perceived behavioural control, such that individuals with higher levels of psychological symptoms and greater reliance on maladaptive coping strategies would exhibit poorer adherence. Finally, it was predicted that the intention-behaviour gap would be primarily attributable to the failure to translate positive intentions into strict adherence, and that this group would exhibit a more symptomatic psychological profile and higher reliance on maladaptive coping strategies than the inclined actors. Method The participants were 390 individuals (mean age = 44.2; SD = 12.7) who responded to a recruitment email sent to a randomly selected sample of 2989 members of the NSW Coeliac Society, screened as meeting inclusion criteria (biopsy-confirmed coeliac disease, gluten free diet duration >3 months, >18 years). Data was submitted anonymously via a link to the online survey, which took approximately 20-30 minutes to complete. The majority of the sample was female (82.8%), which is consistent with the gender distribution of coeliac disease (2-3:1) (Green & Cellier, 2007), and the gender split within the Coeliac Society database (80% female) (Coeliac Society of NSW, 2010). The mean age at diagnosis was 37.4 years (SD = 12.8) and participants had been on a gluten free diet for a mean of 6.8 years (SD = 7.2). The mean duration of symptoms prior to diagnosis was 9.9 years (SD = 12.5; range 0 – 60 years) with most participants Gluten free diet adherence       8     reporting some symptom recurrence when consuming gluten since diagnosis (15.9% mild; 28.2% moderate; 30.3% severe; 20.5% unsure; 5.1% none). Participants completed measures of demographic (e.g., age, gender, education); and disease factors (e.g., age of diagnosis, duration of gluten free diet, symptoms); as well as the following: The Coeliac Dietary Adherence Test (Leffler et al., 2009) is a newly validated seven-item questionnaire designed to assess gluten free diet adherence in coeliac disease. It has demonstrated good psychometric properties and correlates highly with dietitian rated estimates of adherence (Leffler et al., 2009). Higher scores indicate poorer adherence. As per recommendations participants were also classified as having excellent or very good adherence (score 7-12); moderate adherence (13-16); or fair to poor adherence (17-35) (Leffler et al., 2009). The Coeliac Disease Theory of Planned Behaviour Questionnaire (Sainsbury & Mullan, 2011) is a 17-item direct measure of the components of the theory of planned behaviour in relation to gluten free diet adherence. As suggested (Ajzen, 2006; Francis et al., 2004), a longer interview-derived questionnaire was previously administered and adequate content validity of the direct items demonstrated (Sainsbury & Mullan, 2011), therefore allowing for the sole administration of the direct items here. The intention (α = 0.68), attitude (α = .68), and PBC (α = 0.81) composites are internally consistent; subjective norm is less so (α = 0.43), although this was congruent with interview data, whereby respondents felt their decision to maintain a strict gluten free diet was not influenced by other people (Sainsbury & Mullan, 2011). All items are rated on a 7-point Likert scale, with composite scores reflecting the weighted sum of the relevant items. Higher scores indicate more positive intentions and attitudes, and higher perceptions of normative pressure and perceived control. Gluten free diet adherence       9     The Depression Anxiety Stress Scale (Lovibond & Lovibond, 1995) is a 21-item questionnaire designed to measure the negative emotional states of depression, anxiety and stress. Each question is rated on a four-point scale from “did not apply to me at all” to “applied to me very much or most of the time,” with subscale scores representing the sum of the relevant items, and higher scores indicating more severe or frequent symptoms. Clinical cut-off scores denoting normal, mild, moderate, severe, and extremely severe symptomatology are also provided. The questionnaire has good reliability and validity, and distinguishes well between clinical and community samples (Antony, Bieling, Cox, Enns, & Swinson, 1998). The Eating Disorder Inventory-3 Eating Disorder Risk Scale (Garner, 2004) is a 25-item scale designed to measure the risk of an individual developing an eating disorder based on excessive concerns about dieting, body weight, and problematic eating behaviours. It has excellent psychometric characteristics (Garner, 2004). Normative data and instructions for transforming raw scores are provided for females aged 13-53 years with eating disorders in both inpatient and outpatient settings. Due to the difficulty of performing this transformation on a diverse sample (female and male, non-clinical, 18-65 years), raw scores were summed to provide a total score, which was used in subsequent analyses. Comparisons to population norms are thus not possible; however, the measurement was deemed valid in this context as higher scores still indicate greater risk and the analysis was not concerned with identifying eating disorders per se. The Coping Inventory for Stressful Situations (Endler & Parker, 1999) is a 21-item questionnaire designed to assess three distinct coping styles: task-oriented coping (e.g., problem solving or planning a course of action), emotion-oriented coping (e.g., feeling anxious or blaming oneself), and avoidance coping (e.g., removing oneself from the situation). Each question asks the individual to rate how much they engage in the particular coping behaviour when faced with a