Compared to general population estimates, mood and anxiety disorders are highly prevalent and comorbid with one another among persons with HIV (Bing et al. 2001). While the study of depressive symptoms and disorders among those living with HIV has received significant theoretical and empirical attention (Ickovics et al. 2001; Leserman 2008), linking it with worse adherence to treatment and HIV outcomes, less work has been conducted examining the role of anxiety disorders in general, and panic disorder in particular.

Despite high rates panic disorder among those with HIV, there is little empirical investigation focused on how panic disorder might affect the management of HIV. Panic disorder (PD) is characterized by the misinterpretation of benign physical and cognitive sensations as serious and life-threatening (i.e., elevated heart rate, increased respiratory rate, loss of control), as well as anxious apprehension surrounding the possibility of future panic episodes (Barlow 2002). Among the general population, PD is responsible for substantial functional impairment (Kessler et al. 2006), poor quality of life (Barrera and Norton 2009), and reductions in perceived health (Gadermann et al. 2012). There are at least two clinically relevant reasons for studying PD among those with HIV. First, evidence suggests PD is more prevalent among those with compared to those without HIV (11–16 vs. 4–7 %, respectively; (Bing et al. 2001; Kessler et al. 2006). Second, persons with HIV experience common bodily symptoms as a result of the disease and medications used to treat HIV (Justice et al. 2001) and these symptoms significantly overlap with those of PD (e.g., dizziness, nausea) and may contribute to a “fear of fear” cycle present in PD.

PD may also play a substantial and clinically meaningful role in Antiretroviral Therapy (ART) adherence difficulties among those with HIV (Tsao et al. 2004). There is evidence suggesting a number of medical conditions are associated with high rates of PD compared to the general population including asthma (Goodwin and Eaton 2003), chronic obstructive pulmonary disease (Willgoss and Yohannes 2013), migraine (Smitherman et al. 2013), and spinal pain (Von Korff et al. 2005). Although rates of PD may be higher in certain medical conditions, little is understood as to how PD may specifically impact HIV disease management. In a nationally representative probability sample, PD was associated with greater odds (OR = 2.0) of non-adherence than non-psychiatric controls (Tucker et al. 2003). In a nationally representative sample of HIV-positive adults, PD was related to the experience of pain, after controlling for HIV disease stage (Tsao et al. 2004). It is important to further evaluate potential associations between PD and HIV symptom burden and ART adherence.

Adherence to antiretroviral therapy (ART) is a vital component of successful viral suppression, delays in clinical progression, and decreased AIDS-related deaths (Emamzadeh-Fard et al. 2012; Garcia de Olalla et al. 2002; Low-Beer et al. 2000; Paterson et al. 2000). In addition, high ART adherence can decrease the risk for developing resistance to ART (Gardner et al. 2008; Kuritzkes 2004). Moreover, ART adherence may have important secondary HIV prevention implications: effective use of ART and attaining a suppressed viral load may reduce the likelihood of transmission of the virus (Cohen et al. 2011; Del Romero et al. 2010; Donnell et al. 2010; Fisher et al. 2010).

The current study sought to evaluate PD with respect to HIV symptom distress and HIV medication adherence among patients with HIV and symptoms of depression in treatment for opioid dependence. The sample utilized for the current study is useful because the multimorbid psychiatric presentation is unfortunately common among persons living with HIV. Of note, HIV-positive drug users represent an already at-risk population in terms of poor medication adherence (Gonzalez et al. 2011, 2013; Justice et al. 2001) and individuals with substance dependence, including opioid dependence may be at elevated risk for PD (Grant et al. 2004; Martins et al. 2009). Additionally, opioid dependence has been specifically associated with poor ART adherence (Altice et al. 2010). It was hypothesized that those with a current diagnosis of PD, compared to those without current PD, would report greater HIV symptom distress and lower medication adherence, above and beyond potentially relevant demographic and clinical factors.

This is the first study, of which we are aware, showing that current PD diagnosis is associated with greater HIV symptom distress and reduced ART medication adherence in adults with HIV managing opioid dependence. These associations were observed after accounting for a current major depression diagnosis, as well as two other prevalent anxiety disorders: social anxiety disorder and generalized anxiety disorder.

Elevated rates of PD among individuals with HIV (Bing et al. 2001) and those with opioid dependence have been documented (Martins et al. 2009). In line with past findings, the prevalence of current PD was markedly high in this sample, with 29 % of participants positively assessed for a current, primary diagnosis of PD; a figure approximately tenfold higher than that documented in the general population (Kessler et al. 2005a). Negative effects of PD on social and occupational functioning have been established in the general population (Kessler et al. 2005b). However, little is known about the effects of PD on living with and managing HIV and opioid dependence. The current results build on this previous work (Gonzalez et al. 2012) and indicate a potentially unique role for PD in HIV symptom distress and medication adherence in this multiply challenged sample.

Only recently has research begun to explicate potential cognitive mechanisms between PD and HIV. Cognitive factors implicated in PD such as anxiety sensitivity may play a contributing role in the observed relationships. Anxiety sensitivity, conceptualized as the fear of anxiety-related sensations, is an established cognitive risk and maintenance factor for PD (Ehlers 1995) and is associated with HIV-symptom distress (Gonzalez et al. 2012). In addition, Gonzalez et al. (2012) found that HIV symptom distress is related to PD symptoms in the context of elevated anxiety sensitivity. Specifically, side effects of HIV and medications used to treat the virus commonly overlap with panic symptoms. Those with PD or elevated AS may be particularly at risk for misinterpreting HIV symptoms or symptoms of opioid withdrawal as anxiety provoking, resulting in both increased anxiety-related symptom distress and poor medication adherence in an effort to avoid distressing medication-related side effects.

Although limited work has documented elevated rates of PD among individuals with opioid dependence (Gros et al. 2013), research is needed to examine potential underlying mechanisms of this relation, such as the role of acute opioid withdrawal and cognitive vulnerability factors (i.e., AS) as noted previously. Initial work supports a relation between opioid use and elevated AS (Lejuez et al. 2006). Of note, the majority of individuals in the current sample were receiving treatment(s) to maintain abstinence, which may reduce the impact of acute opioid withdrawal symptoms on observed associations. In addition, we have made efforts to account for variables of theoretical and empirical relevance in the models presented but it is plausible that additional factors unique to opioid dependence treatment may be contributing to the observed relationships (e.g., side effects of methadone treatment). Thus, the current findings should be viewed within this context. Future work would benefit from assessing observed relationships at multiple times during the course of opioid treatment (i.e., initiation, dropout).

Further investigations are warranted to parse apart observed relationships and elucidate directionality of these findings. Specifically, the analyses presented here are cross-sectional in nature; therefore neither causality nor directionality of the relationships can be determined. Future research would benefit from examining these associations longitudinally, by manipulating the variables of interest (i.e., treating PD) to evaluate the potential secondary effects on HIV specific distress and adherence or as mediators of treatment outcome, and assessing panic attack history and frequency. Although we attempted to account for the potential influence of major depressive disorder, the effects of major depression or subsyndromal depression cannot be completely ignored. In particular, depression severity and history was not explored.

The measure of HIV symptom distress was modified for these analyses to reduce the risk of construct overlap with PD but resulted in lower internal consistency relative to previous studies (Marc et al. 2012). Lastly, PD accounted for a statistically significant effect, though the portion of overall variance in HIV symptom distress and medication adherence was modest. It will be important to replicate analyses in broader samples of HIV-positive adults, including those not managing opioid use and to investigate other factors that might reduce adherence (e.g., neurocognitive deficits; Zogg et al. 2012).

The results presented here provide incremental evidence for specific and clinically important effects of PD in individuals with multiple competing and chronic psychological and physical health conditions. These results underscore the impact and importance of assessing for and treating PD in persons with HIV managing opioid use and point to the importance of developing flexible treatment platforms capable of addressing multiple comorbid mental health problems in this patient group.