Predictors of anxiety in multiple sclerosis.

Multiple sclerosis (MS) is a chronic neurologic condition that impacts as many as 2.3 million individuals worldwide (“Atlas of MS Database,” 2013). The disease course and symptom profile varies significantly between individuals and usually progressively worsens over time. Typically, patients experience multiple symptoms resulting in physical, cognitive and psychological difficulties. Although a wealth of research has focused on depression and its treatments in MS, little is known about symptoms of anxiety and risk factors for its development in MS. Hence, currently there exists no unified theoretical conceptualization of the experience of anxiety in MS and how that may change over time. This is a surprising gap in the literature, especially given that prevalence of anxiety disorders in MS ranges from 14% to 45% (Korostil & Feinstein, 2007; Wood et al., 2013) and may result in poorer medication adherence (Turner, Williams, Sloan, & Haselkorn, 2009), higher pain intensity and pain interference with health- related quality of life (Bruce & Arnett, 2009; Kalia & O’Connor, 2005), lower quality of life (Garfield & Lincoln, 2012), and suicidal intent (Korostil & Feinstein, 2007). Identification of risk factors for anxiety in MS would lay a critical foundation for developing an advanced conceptualization of the experience of anxiety in MS over time. This holds tremendous practical implications in that such a theoretical advancement would enhance clinical efforts directed towards the prevention, screening, and treatment of anxiety and concomitant secondary symptoms. In the absence of prior risk factor research for anxiety in MS, it is informative to consider if the cross-sectional variables associated with anxiety in MS may also be predictive of symptoms over time. Higher anxiety has been associated with female gender (Jones et al., 2012), younger age (Wood et al., 2013), a diagnosis of relapsing-remitting MS (Jones et al., 2012), longer time since onset of MS (though results are mixed; Feinstein, O’Connor, Gray, & Feinstein, 1999; Janssens et al., 2006), greater MS-related disability (Jones et al., 2012), higher levels of fatigue (Beiske et al., 2008), MS exacerbations (rapid onset or an increase in symptoms; Burns, Nawacki, Siddique, Pelletier, & Mohr, 2013; Feinstein et al., 1999), and depression (Garfield & Lincoln, 2012; Wood et al., 2013). Investigation is needed to identify if these cross-sectional correlates may represent predictive risk factors that have an extended impact on the development, maintenance, or exacerbation of anxiety symptoms over time. A challenge for identifying risk factors of anxiety is the high comorbidity between anxiety and depression (Almeida et al., 2012; Burns, Siddique, Fokuo, & Mohr, 2010). Evidence suggests that individuals with a history of depression may be at a higher risk of experiencing anxiety over time (Korostil & Feinstein, 2007). A further challenge is that anxiety, depression, and MS include overlapping somatic symptoms (e.g., sleep disturbance, fatigue, concentration difficulties). Similarly, reports of ‘numbness and tingling,’ ‘feeling unsteady,’ and ‘wobbliness in legs’ might be attributed to anxiousness or the MS disease process (Feinstein et al., 1999). A number of studies have discussed the importance of removing somatic symptoms from both depression and anxiety scales to avoid this confound (Benedict, Fishman, McClellan, Bakshi, & Weinstock-Guttman, 2003; Donnchadha et al., 2013). Recommendations include utilizing an outcome measure comprised only of non-somatic items (i.e., excessive worry, fear of losing control, unable to relax, etc.) of anxiety (Donnchadha et al., 2013; Zigmond & Snaith, 1983) or utilizing a two-factor model of somatic vs. non-somatic depression symptoms (Richardson & Richards, 2008). The present study sought to extend the literature on anxiety in MS by conducting a evaluation of current demographic, disease-related characteristics and psychological factors that contribute to anxiety symptom severity at a 4-month follow-up. In the absence of prior longitudinal research on anxiety among the MS population, we hypothesized that variables cross-sectionally associated with anxiety would also remain predictors during 4-month follow-up: age, level of disability, time since onset, and depression. To account for the concerns of somatic overlap, we utilized an anxiety measure that focuses on cognitions (and excludes somatic items), and a two-factor measure of depression that differentiates somatic and non-somatic symptoms. We hypothesized that the non-somatic depression symptoms would be associated with anxiety symptom severity due to the substantial overlap between the affective/cognitive symptoms of anxiety and depression derived from the shared distress and negative affect (Kendler et al., 1995; Watson & Clark, 1984; Watson, Clark, & Carey, 1988). Finally, since there is considerable reason to believe that both depression and anxiety may be influenced by age (Janssens et al., 2006; Jones et al., 2012; Mattioli, Bellomi, Stampatori, Parrinello, & Capra, 2011; Williams et al., 2005; Wood et al., 2013), time since onset (Chwastiak et al., 2002; Forman & Lincoln, 2010; Janssens et al., 2006; Korostil & Feinstein, 2007; Williams et al., 2005), and disability (Beiske et al., 2008; Mattioli et al., 2011; Moore et al., 2012), we explored the potential synergistic (i.e., interaction) effect between depression and age, time since onset, and disability on anxiety over time to determine if a differential pattern of association would emerge between these factors.