ATTENTION-DEFICIT/HYPERACTIVITY DISORDER, STRESS, AND QUALITY OF LIFE IN ADULTHOOD A Thesis By MARTHA

The symptoms of Attention-Deficit/Hyperactivity Disorder (ADHD) are linked to dysfunction in academic, occupational, interpersonal, and other domains for both children and adults (see review in Whalen, Jamner, Henker, Delfino, & Lozano, 2002). The current study examines the association between ADHD symptoms and quality of life (QOL) and perceived stress in a community sample of adults. Data were obtained largely through an internet survey conducted as part of a larger medical trial utilizing a representative community sample, and were analyzed through a series of hierarchical multiple regressions employing ADHD symptom clusters, demographic, and anxiety and depression scale variables as predictors. The hypotheses that ADHD symptoms positively predict perceived stress and negatively predict QOL were generally supported, with inattention and sluggish cognitive tempo serving as the strongest ADHD-related predictors for both QOL and perceived stress. ADHD, Stress, and Quality of Life 3 Attention-Deficit/Hyperactivity Disorder, Stress, and Quality of Life in Adulthood In the past, it was commonly believed that Attention-Deficit/Hyperactivity Disorder (ADHD) affected only children and adolescents. Empirical research, however, has shown ADHD to be a life-long disorder, with ADHD traits continuing into adulthood in approximately 60% of diagnosed children and adolescents (Sobanski, Schredl, Kettler, & Alm, 2008). Some have even suggested that the prevalence rates of ADHD in childhood and adulthood are comparable, with 3-7% of children (American Psychiatric Association, APA, 2000) and 4.4% of adults (Kessler et al., 2006) demonstrating adequate symptoms to merit an ADHD diagnosis. ADHD is primarily characterized by two symptom clusters, inattention (IA) and hyperactivity-impulsivity (HI), and is divided into types based on expressed symptoms in these clusters. The two most prevalent types are Primarily Inattentive (ADHD-IA) and Combined (i.e., prominent HI and IA symptoms), with the third subtype, Primarily Hyperactive (ADHD-HI), being much less prevalent and potentially developmentally limited to childhood (see below). IA traits are characterized by a variety of behaviors, including difficulty concentrating, a tendency to daydream or become easily distracted, and lessened ability to organize and plan (Barkley, 2006). On the other hand, it has been suggested that the primary deficit of individuals with prominent HI symptoms is behavioral response inhibition (Barkley, 1997). Individuals with elevated HI tend to respond without waiting for directions to be completed or considering possible consequences (i.e., are more prone to engage in risky behaviors; Barkley, 2006). Other activities that require sharing, cooperation, and impulse control pose challenges for these individuals, as well (Barkley, 2006). Prevalence rates differ ADHD, Stress, and Quality of Life 4 between ADHD types. In the United States, the prevalence of ADHD-IA in children and adolescents has been estimated between 5.4 and 8.8% while approximately 2.6% display primarily hyperactive/impulsive traits. Individuals with symptoms of both inattention and hyperactivity/impulsivity make up the combined type (ADHD-C) of ADHD. The prevalence rate for ADHD-C in childhood is estimated at 4.7% (see review in Barkley, 2006). Some research has shown that the diagnosis of a particular ADHD subtype tends to be associated with the age of the individual. One study by Nolan, Volpe, Gadow, and Sprafkin (1999) examined subtype prevalence in children and adolescents (ages 3 to 18) that had been referred to a child psychiatry outpatient clinic for ADHD symptoms. Sixty-three percent of the sample was diagnosed by a psychiatrist as meeting criteria for one of the ADHD types. The remaining 37% exhibited sub-clinical ADHD symptoms. Results from this study showed the mean age of individuals classified as ADHD-IA to be 11.98 (SD = 3.48), significantly higher than the mean age for the ADHD-HI group (M = 7.70, SD = 4.11) or the ADHD-C group (M = 8.63, SD = 4.35; Nolan et al., 1999). This would suggest HI to be more common in younger individuals, while inattention may be a more enduring symptom. Research has traditionally indicated the prevalence of ADHD in males to be above that of females (see review in Barkley, 2006). Research by Nolan and colleagues (1999) found the ratio of ADHD in male children compared to female children to be similarly high across all three types (IA: 76% vs. 24%; HI: 81% vs. 19%; and C: 78% vs. 22%). As evidence for the continuation of ADHD into adulthood accumulates, differences between the child and adult ADHD populations become clearer. Research has shown that, indeed, as children with ADHD-HI symptoms mature, these symptoms become less evident than they were in childhood (see review in Barkley, 2006). Although cognitive ADHD, Stress, and Quality of Life 5 manifestations of impulsivity (e.g., poor time management and a sense of restlessness) may still exist for the ADHD adult, the physical over-activity appears to be less evident (APA, 2000). Due to the fact that ADHD manifests somewhat differently in adults than it does in children, for whom the DSM-IV-TR criteria were written, assessment of the disorder in older individuals poses certain challenges. For instance, inattention is a common, if sporadic, event for most adults; determination of what constitutes “abnormal” inattention can be subjective. Additionally, the diagnosis of ADHD in adults is dependent on self report of historical childhood symptoms, which invites risk of inaccuracy or incompleteness (Barkley, 2006). For these reasons as well as others, attention has been paid to the development of diagnostic criteria appropriate for adults demonstrating ADHD symptoms (Barkley, 2006). In fact, Russell Barkley, a prominent ADHD researcher, has proposed new criteria for the diagnosis of ADHD in adults, taking into account the differences evident in adulthood (Barkley, Murphy, & Fischer, 2008). The trait of sluggish cognitive tempo (SCT), characterized by sluggishness, passivity, confusion, and hypoactivity has been linked to ADHD-IA in children (Barkley, 2006; Carlson & Mann, 2002). These symptoms appear to oppose those associated with ADHD-HI and, due partly to the fact that HI is not as prevalent in the adult population, SCT is now considered important to assess as a predictor of ADHD, and particularly the Inattentive type, in adults. Symptoms associated with all three types of ADHD are generally associated with pervasive difficulties in academic, occupational, interpersonal, and emotional domains (Whalen, Jamner, Henker, Delfino, & Lozano, 2002). Long-term deficits in these areas seem very likely to affect a person’s quality of life and overall level of perceived stress. ADHD, Stress, and Quality of Life 6 Academic Difficulties Associated with ADHD Academic difficulty for those diagnosed with ADHD can begin at an early age, and is common. One study showed 68% of children in an ADHD group to have learning disabilities or academic problems, whereas only 21% of a non-diagnosed control group demonstrated such deficits (Loe et al., 2008). Further, while the intellectual ability of persons with ADHD has generally been shown to be comparable to those without the disorder (Barkley, 2006), a study by Loe and colleagues (2008) demonstrated that a sample of children with ADHD, who had also been referred for a developmental evaluation, performed poorly on the WISC-III (Weschler Intelligence Scale for Children), as compared to non-diagnosed peers (mean FSIQ 88.98 +/14.82, versus 101.58 +/14.01; Loe et al., 2008). While the data are inconclusive regarding overall cognitive ability of typical students with ADHD, it is incontrovertible that many struggle academically (Barkley, 2006; DuPaul, 1991; McDonald-Richard, 1995; Reardon & Nagliery, 1992; Reaser, Prevatt, Petscher, & Proctor, 2007). In fact, Biederman and colleagues (1998) concluded that up to 84% of children with ADHD need academic assistance beyond the level of that given to nondiagnosed peers. Factors accounting for this population’s academic difficulties are abundant, but inattention seems directly linked to the most common problems. These include attending to relevant classroom and academic information, executing necessary learning strategies such as planning and organizing (Reardon & Nagliery, 1992), implementing self-regulatory behaviors in academic settings, and persisting with tasks until completion (McDonaldRichard, 1995). Gaub and Carlson (1997) explored behavioral, academic, and social differences among children with ADHD-IA, ADHD-HI, ADHD-C, and a non-diagnosed control group ADHD, Stress, and Quality of Life 7 (NC), based on teacher reports. While the ADHD-IA group garnered significantly higher appropriate behavior ratings than the other two diagnosed groups, it, along with the ADHD-C group, scored significantly lower than the ADHD-HI group on overall learning abilities. These results imply that learning difficulties and academic problems are less common in solely hyperactive-impulsive types, suggesting that, among the ADHD symptom clusters, inattention may be a more influential factor for learning difficulties. While inattention may be primary in terms of effects on learning, hyperactivity and impulsivity (HI) have also been shown to have negative effects on academic performance. For those who are diagnosed with ADHD in the pre-school years, HI tends to be more present than inattention (Barkley, 2006), and the presence of HI at such an early stage generally predicts chronic behavior problems (Farrington, Loeber, & Van Kammen, 1990; Sonuga-Barke, Auerbach, Campbell, Daley, & Thompson, 2005), deficits in academic fluency (i.e., reading disorder) and, in some cases, intellectual impairment (DuPaul, McGoey, Eckert, & VanBrakle, 2001; McGee, Partridge, Williams, & Silva, 1991). Those with significant HI later demonstrate increased truancy, grade retention, and dropout (Barbaresi, Katusic, Colligan, Weaver, & Jacobsen, 2007) and other behavioral problems at school (Loe et al., 2008). In summary, inattention, hyperactivity, impulsivity, or a combination of these can negatively impact the general ability to perform the skills needed to succeed academically (Barkley, 2006). Reaser and colleagues (2007) compared the learning strategies of college students with ADHD to peers with learning disabilities (LD) and another group without ADHD or LD, finding that the ADHD group demonstrated significantly lower achievement in the areas of main idea selection and test-taking strategies than either of the other two groups. ADHD, Stress, and Quality of Life 8 Additionally, the ADHD group performed lower than the non-diagnosed group in the areas of concentration (d = .40), motivation (d = 1.20), and information processing (d = .61; Reaser et al., 2007). These results reinforce the enduring negative impact that inattention, hyperactivity, and impulsivity can continue to have on academic functioning into adulthood. Occupational Problems Associated with ADHD Many of the deficits that affect those with ADHD academically, such as impaired attention, interpersonal skills, initiative, and task persistence, also hinder occupational functioning (Hill & Petty, 1995), which becomes increasingly evident as ADHD adolescents grow into adults and take on jobs requiring more skill and responsibility (Barkley, 2006). Not surprisingly, academic impairment has been demonstrated as a potential reason for later occupational dysfunction in those with ADHD. One study found that 32% of an ADHD sample had failed to complete high school, a finding that helps to explain a trend for ADHD adults to have both lower-status occupations (Weiss, Hechtman, Milroy, & Perlman, 1993) and lower general SES (Barkley, 2006) and supports the link between education level, occupational class, and income (Laaksonen, Rahkonen, Martikainen, & Lahelma, 2005). Evidence for occupational impairment in adults with a diagnosis of ADHD is varied and includes higher rates of being fired, laid off (Barkley, Fischer, Smallish, & Fletcher, 2006), and moving or changing jobs (Weiss et al., 1993), when compared to those without ADHD. Able, Johnston, Adler, and Swindle (2007) illustrate this trend, reporting adults with an ADHD diagnosis to have lower levels of full-time employment (70% vs. 88.6%) and higher frequency of unemployment (11% vs. 2.8%) than a group of non-symptomatic peers. Additional evidence for occupational impairment is found in long-term follow-up research by ADHD, Stress, and Quality of Life 9 Barkley and colleagues (2006) that reports significantly lower employer-rated job performance for adults with ADHD as compared to a non-diagnosed control group. Interpersonal Difficulties Associated with ADHD A large body of evidence suggests that ADHD children, regardless of type, experience more social difficulties than their non-diagnosed peers (e.g., Gaub & Carlson, 1997; see review in Canu & Carlson, 2004). Several researchers (e.g., Wheeler & Carlson, 1994; see review in Henker & Whalen, 1999) describe a tendency for children with the different ADHD types to display varying social dysfunctions; parents and teachers characterize ADHD-IA children as passive, shy, and socially withdrawn, whereas they describe ADHD-C children as inappropriately assertive, engaged to the point of intrusiveness, and emotionally labile (Maedgen & Carlson, 2000). When compared to their non-diagnosed peers, children with both types of ADHD experience increased levels of anxiety and depression associated with their social difficulties (Kellner, Houghton, & Douglas, 2003). Other research by Gaub and Carlson (1997) used teacher reports to assess the social functioning of children with all types of ADHD (-IA, -C, and -HI). Results showed all three groups as demonstrating significantly more social deficits than children without ADHD, with the ADHD-C group demonstrating the most. In addition, different levels of “peer dislike” were detected for each of the ADHD subgroups: fifty-nine percent of ADHDIA and 53% and ADHD-HI group members were disliked by their peers, whereas the ADHD-C group had a higher rate of 82% (Gaub & Carlson, 1997), suggesting that when a significant number of both inattentive and hyperactivity-impulsive symptoms are present, social adjustment suffers more. ADHD, Stress, and Quality of Life 10 Empirical research has also suggested that children with ADHD have fewer friendships than children without the disorder (Erhardt & Hinshaw, 1994; Hoza et al., 2005; see review in Canu & Carlson, 2004). Male summer campers—regardless of ADHD status— tend to describe other male campers with ADHD as more negative and less positive than non-diagnosed peers after only three days of interaction, an effect that was replicated at the end of the five-week program (Erhardt & Hinshaw, 1994). Hoza and colleagues (2005) document 56% of ADHD children as having no dyadic friends, 33% with one dyadic friend, and 9% with two dyadic friends, as compared to 32%, 39%, and 22% in those categories for non-diagnosed children. Overall, in this study, ADHD children were described as more frequently rejected (52% vs. 14%) and less popular than other children (Hoza et al., 2005). While ADHD-related social difficulties usually begin in childhood and have been found to continue into adolescence, it is becoming clear that affected adults also suffer relationally (Bagwell, Molina, Pelham, & Hoza, 2001; Barkley, 2006; Weiss & Murray, 2003). Research by Young and Gudjonsson (2006) compared a group of ADHD adults to sub-clinical ADHD and non-symptomatic groups using the Gough Socialization Scale (GSS), a measure of social adjustment difficulties, including struggles with forming and maintaining intimate relationships. Results showed the ADHD group to score similarly on the GSS to people meeting criteria for a personality disorder, equating to a substantially increased likelihood for social, relational, and adjustment problems. It is possible that some social deficits experienced by those with ADHD may be the result of negative peer evaluation based on stigma toward ADHD as a mental disorder (Canu, Newman, Morrow, & Pope, 2008). Research has shown that college students desire less interaction with hypothetical peers with ADHD, across levels of personal intimacy, as ADHD, Stress, and Quality of Life 11 compared to peers with no differences besides lacking an ADHD diagnosis (Canu et al., 2008), which suggests that an adult ADHD label carries with it negative stigma. Adults with ADHD are also likely to struggle interpersonally as a result of their continued inattention, hyperactivity-impulsivity, or both (Robin & Payson, 2002). In a study by Canu and Carlson (2003), interpersonal deficits were manifest in comparisons of male college students with ADHD-IA, ADHD-C, and no ADHD diagnosis. Those with ADHD-IA reported significantly lower levels of social comfort, assertiveness, and frequency of dating behaviors, as compared to the other two groups. Interestingly, in terms of heterosexual romantic outcomes, the ADHD-C group’s adjustment seemed more or less indistinguishable from the control group. Emphasizing the social impairment of young men with ADHD-IA is the finding that, when blind to ADHD status, female confederates and observers reacted more negatively to males with ADHD-IA, across several measures, as compared to their reactions to those in the ADHD-C and control groups (Canu & Carlson, 2003). It has also been suggested that overactive perception of negative social evaluation combined with low social comfort (i.e., higher rejection sensitivity) may differentially and negatively affect individuals with ADHD-IA (versus those with ADHD-C), increasing the likelihood of their experiencing social difficulties and peer rejection (Canu & Carlson, 2007). Finally, elevated rates of relationship instability, spousal separation, and divorce have also been noted in the population of adults with ADHD (Barkley, 2006; Murphy & Barkley, 1996; Weiss & Murray, 2003). Emotional Difficulties Associated with ADHD Ample research has established that the problems associated with ADHD can lead to personal distress. In addition, those with ADHD also frequently meet criteria for a comorbid ADHD, Stress, and Quality of Life 12 Axis I disorder. The majority of research on ADHD comorbidity has focused on conduct disorder (CD), oppositional defiant disorder (ODD), and other childhood aggression. Far less attention has been paid to the mood and anxiety disorders (Jensen, Martin, & Cantwell, 1997), despite their high co-occurrence with ADHD. Overall, anxiety disorders are present in up to 25% of children with ADHD, and clinical anxiety seems overrepresented in the adult ADHD population as well (Barkley, 2006). A study by Bohline (1985) found that teachers rated ADHD children as demonstrating more depressive features than non-diagnosed children, whereas research by Waring and Lapane (2008) found that when compared to their non-diagnosed peers, children with ADHD were approximately nine times as likely to experience significant depression, anxiety, or both. Research also suggests that the presence of CD and ODD with ADHD in childhood may serve as a predisposing factor for the development of anxiety and mood disorders later in life (Bagwell, Molina, Kashdan, Pelham, & Hoza, 2006; Barkley, 2006). Indeed, a study by Bagwell and colleagues (2001) comparing children with and without ADHD demonstrated that social anxiety disorder occurs in children who had been diagnosed with ADHD and comorbid CD or ODD at a rate of more than 2.5 times that of their nondiagnosed peers (Bagwell et al., 2001). Regarding anxiety and depression comorbidity beyond adolescence, adults with ADHD, as compared to sub-clinical controls and non-symptomatic controls, report substantially more anxious and depressive experiences (Young & Gudjonsson, 2006). This trend is supported by Able and colleagues (2007), who examined a sample of adults without ADHD, with undiagnosed ADHD, and with a documented diagnosis of ADHD, and demonstrated that the undiagnosed ADHD group was more likely than non-diagnosed ADHD, Stress, and Quality of Life 13 controls to have experienced self-reported depression (31% vs. 12.9%), anxiety (19.8% vs. 7.3%), and bipolar disorder (5.9% vs. 0.9%). Additionally, the diagnosed ADHD group reported incrementally higher rates of these dysphoric moods as compared to their peers with undiagnosed ADHD (55.2% for depression, 30.1% for anxiety, and 10.3% for bipolar disorder). Considering many of the previously mentioned difficulties (e.g., academic, interpersonal difficulties) and their emotional toll (i.e., lower motivation, fewer friends), the adult with ADHD’s heightened vulnerability for the development and maintenance of anxiety and other emotional problems becomes even clearer. ADHD and Stress Defining stress. Stress has typically been viewed as a negative emotional experience with the capacity to affect psychological and physical health (Taylor, 2006). When presented with a stressor, an individual experiences physiological, cognitive, and behavioral changes that facilitate coping and problem solving. However, the construct of stress is subjective, and is manifest differently in different people. Symptomatic responses to a potentially stressful stimulus differ among individuals, whose various appraisals of those symptoms as positive or negative contribute to the experience of stress (Taylor, 2006). The personal stress experience, therefore, can be physiological, emotional, or a combination of the two. ADHD and stressor prevalence. Given information about the host of difficulties with which people with ADHD must deal (e.g., academic struggles, interpersonal rejection, comorbid mood or anxiety issues, as described above) it is logical to assume that the ADHD individual experiences increased levels of subjective, day-to-day stress. The impact of higher stress is likely not limited to the sufferer, but also probably affects family, friends, and one’s personal environment, further supporting the assumption that ADHD individuals have a ADHD, Stress, and Quality of Life 14 higher number of stressors in their lives. ADHD exacts numerous costs (i.e., behavioral interventions, medication costs, time off from work; Riley, Lyman et al., 2006) upon families of diagnosed children and affected adults, alike, and the stress associated with such costs ultimately affects the environment (e.g., family living situation; Beitchman, Inglis, & Schachter, 1992; Lange et al., 2005). Research has shown that close to $3.7 billion was lost in 2000 in relation to ADHD, whether due to unemployment of family members who miss work due to problems associated with their children having ADHD or work loss among those with a diagnosis of ADHD (Birnbaum et al., 2005), illustrating the magnitude of economic stress incurred within a family. To the extent that the individual with ADHD realizes, or is made aware of, his or her contribution to family and general life stressors, personal subjective stress experience may also intensify. Stress reactivity and ADHD. Direct research regarding the ADHD-personal stress association has focused mainly on physiological stress responses rather than on extent of subjective daily stressors. The physiological stress response and stress experience in adult ADHD patients has been examined by comparing the event-mediated arousal of adults with ADHD to that of non-diagnosed adults. In a study by Lackschewitz, Huther, and KronerHerwig (2008), the Trier Social Stress Tasks (comprised of a five minute simulated job interview and a five minute mental arithmetic task in front of a committee) were administered while heart rate and salivary cortisol levels were being measured. Following the task, the participants completed a questionnaire to explore the subjective stress experience. Results indicated elevated physiological stress responses and higher subjective stress among the ADHD adults when compared to the control group. Additional research supports this finding, suggesting dysregulation of the stress-mediating mechanisms (Randazzo, Dockray, ADHD, Stress, and Quality of Life 15 & Susman, 2008) is related to increased stress for those with pervasive ADHD symptoms (King, Barkley, & Barett, 1998). In fact, the levels of reported stress for those adults diagnosed with ADHD were elevated both in the presence of the stressor and in anticipation of it. Additionally, measures of heart rate and salivary cortisol demonstrated a greater difficulty in recovering from elevated stress levels in the ADHD group (Lackschewitz et al., 2008), indicating that persons with ADHD not only have a more severe physiological stress experience, but also face challenges in recovering from stress. These data provide evidence for the natural predisposition of an ADHD individual to higher stress reactivity. Stress management skills and ADHD. Over a decade ago, inability to tolerate stress was proposed by Wender as an additional characteristic for the diagnosis of ADHD (Wender, 1995). Indeed, research has suggested that individuals with ADHD demonstrate lower coping skills in the face of stress (Riley, Spiel et al., 2006), a condition that could exacerbate the impact of stressors. Children with ADHD have been found to demonstrate lower self-esteem than non-diagnosed peers and, as a result, engage is less active approaches to solving interpersonal problems (Riley, Spiel et al., 2006). The symptoms of inattention, hyperactivity, and impulsivity are linked to lower levels of planning, organization, and selfcontrol (Davis, Levitan, Smith, Tweed, & Curtis, 2006), which are critical to successful stress-management. Individuals with ADHD, especially those with primarily hyperactive or impulsive traits, tend to act immediately when presented with a complex stimulus rather than to engage in the often necessary problem-solving processes (Davis et al., 2006). If the stimulus is considered by the individual to be stressful, his or her impulsive response likely will not be well planned, and therefore is unlikely to actually alleviate stress. ADHD, Stress, and Quality of Life 16 ADHD and Quality of Life The Centers for Disease Control and Prevention (CDC, 2009b) describes healthrelated quality of life (QOL) to be a personal perception of physical and mental health and well-being throughout the lifespan. Health-related quality of life has been incorporated into several areas of research, including as a measure of the effects of disorders. QOL has broadly been used to refer to many aspects of health and life satisfaction (CDC, 2009b). While the literature in this area is not extensive, ADHD has been associated with various measures of impaired QOL (Riley, Spiel et al., 2006). Using parent reports, Riley, Spiel, and colleagues (2006) assessed the QOL of children in five domains: risk avoidance, achievement, satisfaction, resilience, and comfort. Children with ADHD reported low levels of well-being and self-worth (e.g., poor satisfaction), increased frequency of emotional and somatic complaints (e.g., low levels of comfort), decreased involvement in family and fewer coping skills (e.g., poor resilience), increased likelihood to engage in dangerous behaviors (e.g., poor risk avoidance), and poorer academic and social functioning (e.g., poor achievement), as compared to their non-diagnosed peers. All of these results are robust indices of lower QOL (Riley, Spiel et al., 2006). Able and colleagues (2007) later examined the association of ADHD traits with adults’ QOL. Using telephone surveys, the experimenters collected data on several aspects of QOL from adults diagnosed with ADHD, others reporting significant but undiagnosed ADHD symptoms, and peers without the disorder. Results showed that all adults with diagnosable levels of ADHD reported higher frequencies of past and present comorbid conditions, increased likelihood for family history of ADHD, lower levels of educational achievement, lower frequencies of full-time employment, lower socio-economic status, less ADHD, Stress, and Quality of Life 17 stability in interpersonal relationships, increased likelihood for depressive symptoms, and an increased risk for accidents and injuries (Able et al., 2007), which all speak to diminished QOL. The extent of emotional QOL deficit experienced by adolescents with ADHD, who are already in a turbulent period of life, is illustrated by the research of Whalen and colleagues (2002). Self-reports from adolescents with moderate-to-high ADHD symptoms show increased frequency of anxiety, sadness, anger, and stress, as well as decreased happiness and positive well-being, when compared to other peers. Interestingly, despite previous evidence for social deficits, adolescents with high levels of ADHD also reported spending more time with peers and less time with family (Whalen et al., 2002), perhaps losing out on more intimate emotional support. As noted previously, ADHD is a heterogeneous condition, sometimes presenting with comorbidities that may be more obvious or seem more urgent to medical or mental health professionals, and which may mask the underlying ADHD (Able et al., 2007). Anxiety, depression, conduct disorder, and learning disabilities are common comorbidities, as are other environmental stressors (e.g., disordered home or school environment; Able et al., 2007). Such comorbidities exacerbate the already turbulent daily social, emotional, and work experiences of those with ADHD, all of which are implicitly linked to subjective QOL. ADHD has also been linked to problematic choices and behaviors that can have longterm effects on physical health, an additional aspect of QOL. Repeated evidence has been found linking ADHD symptoms to increased frequency of cigarette smoking (Lambert & Hartsough, 1998; Whalen et al., 2002; Whalen, Jamner, Henker, Gehricke, & King, 2003), and alcohol and drug abuse (Murphy & Barkley, 1996), as well as to less healthy diet choices ADHD, Stress, and Quality of Life 18 (Davis et al., 2006), and even obesity in adolescents (Cortese et al., 2007; Lam & Yang, 2007; Waring & Lapane, 2008) and adults (Altfas, 2002). Summary and Rationale for the Current Study A variety of life experiences in childhood and adulthood are negatively impacted by the presence of ADHD symptoms. Regardless of the fact that individuals with ADHD demonstrate comparable levels of intelligence to those without the disorder (Barkley, 2006), academic and occupational difficulty is common (Reaser et al., 2007; Barkley, 2006; Weiss et al., 1993). Cognitive abilities related to academic and occupational functioning, such as concentration, information processing, and motivation, are typically impaired in affected individuals (Reaser et al., 2007), as are interpersonal relationships in childhood (Erhardt & Hinshaw, 1994; Hoza et al., 2005; see review in Canu & Carlson, 2004) and adulthood (Barkley, 2006; Murphy & Barkley, 1996; Weiss & Murray, 2003). Relatedly, individuals with ADHD have a higher prevalence of general emotional problems as well as meeting full criteria for mood disorders (Able et al., 2007; Young & Gudjonsson, 2006), as compared to non-diagnosed peers, and those with ADHD have been shown to be more physiologically reactive to stress (Lackschewitz et al., 2008). QOL of ADHD individuals is nearly certainly affected by dysfunction in all the aforementioned areas, but is also potentially impacted by their elevated likelihoods of smoking, drinking, unhealthy eating, and obesity (Davis et al., 2006; Lam & Yang, 2007; Murphy & Barkley, 1996; Whalen et al., 2002). In sum, the impact of ADHD is significant and plays a considerable role in the ADHD individual’s experience of psychological and physiological well-being. ADHD, Stress, and Quality of Life 19 To date, however, research focusing on the health-and-happiness outcomes associated with ADHD has largely focused on children. With almost 1in 20 adults in the United States affected by ADHD (Kessler et al., 2006), it is clear that this older segment of the ADHD population warrants similar attention. Therefore, this study aimed to examine the impact that ADHD symptoms have on the psychological and physiological health of adults. Specifically, QOL and perceived stress were assessed as dependent variables across an adult, community-recruited sample that is indexed by adult ADHD traits. Based on previous research, this project had clear hypotheses for ADHD’s relation to the aforementioned outcomes in all age groups. A negative relationship was expected between the level of ADHD traits (i.e., IA, HI, Sluggish Cognitive Tempo, and newly proposed adult ADHD criteria) and every index of QOL (physical health, psychological QOL, social relationships, environmental QOL). The comparison of all ADHD traits to level of perceived stress was expected to yield a positive association. Method Participants A sample of over 1,000 non-clinical, community participants from Watauga County in Western North Carolina was recruited via newspaper, radio, and online advertising to participate in a larger study examining the effects of an antioxidant trial on physical and psychological health. An incentive of $300 was offered for completion of the entire study, which involved a three month placebo or antioxidant regimen, baseline and post-trial physiological, cognitive, and online psychological assessments. The sample included males and females from the ages of 18 to 85, and was limited to participants who were noninstitutionalized, not pregnant, and non-lactating. Participants were not excluded on the basis ADHD, Stress, and Quality of Life 20 of ethnicity, gender, or other physical or mental health status. The baseline survey included measures examined in the current study; nine-hundred-eighty-three participants (94.16% of original sample) completed ADHD measures and were included in this study. Sixty-one participants from the original sample were excluded from the current study due to incomplete data. The final study sample included 39.3% male and 60.7% female participants, 95.6% of whom were Caucasian. Age breakdown of the sample is as follows: 37.6% young adult (ages 18-40), 49.9% middle-aged (ages 41-65), and 12.4% at retirement age (ages 66-85). The majority of the sample was highly educated, with 81.6% having completed at least some college. An additional characteristic used to recruit the sample was Body Mass Index (BMI). In this sample, 46.6% of the sample were considered to have a normal BMI (18.5-24.9), 28.6% were overweight (25-29.9), and 24.8% were considered to be obese (30 or more). Additional details regarding the sample’s composition are in Table 1. Procedure Data from the sample of 983 were collected in 2008, with 473 people participating in the spring (cohort one) and 510 people participating in the fall (cohort two). After giving written informed consent, participants completed baseline questionnaires and rating scales including an ADHD assessment scale based on Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; APA, 2000) criteria and criteria recently proposed by Barkley and colleagues (2008) as relevant to adulthood, a QOL measure, and a perceived stress measure. While most of the questionnaires were administered online, spring participants were administered a paper-and-pencil version of the Perceived Stress Scale at the time of the baseline laboratory visit. In addition, a small proportion of all participants did not complete the online survey before presenting at baseline (due to limited internet access or other issues); ADHD, Stress, and Quality of Life 21 these individuals completed the survey in a computer laboratory during the course of their baseline visit. This study’s procedures were reviewed and approved by the Institutional Review Board on May 1, 2009 (see Appendix A for IRB Approval Form and Appendix B for Consent Form for Human Subjects). Measures Current Symptom Scale (Barkley & Murphy, 2006). This self-report measure consists of 18 items tapping the inattentive and hyperactive-impulsive symptoms of ADHD as defined in the DSM-IV-TR (APA, 2000), and references behavior in the past six months. Example items include “Fidgeted with hands or feet or squirmed in seat” (hyperactivityimpulsivity) and “Didn’t listen when spoken to directly” (inattention). Responses are made on a four-point Likert-type scale (0 = never or rarely; 3 = very often). A single item was added to assess overall impairment experienced by participants due to ADHD symptoms, using the same scale. In addition, three items tapping SCT, a cognitive style associated with the Inattentive Type of ADHD (see above for description; Carlson & Mann, 2002), were also added to each measure to examine the potential association of this distinct trait on health outcomes (see Appendix C for Current Symptom Scales). In the current sample, internal reliability for all scales was adequate, as assessed by Cronbach’s alpha, α = .86 (IA), .78 (HI), .73 (SCT). Proposed Adult ADHD Criteria Scale (Barkley et al., 2008). This nine-item selfreport measure has recently been shown in a large adult sample to more effectively distinguish ADHD from non-ADHD subgroups than the DSM-IV-TR ADHD criteria, which were developed using data derived from child and adolescent populations. Three items included in this symptom set are duplicates of DSM-IV-TR criteria; as these were assessed ADHD, Stress, and Quality of Life 22 using the Current Symptom Scale (see above) and included in the IA and HI scales, they were dropped for the purposes of independent analysis, creating a six-item form. Responses are given on a four-point scale similar to the Current and Childhood Symptom Scales (Barkley & Murphy, 2006) and reference the past six months; sample items include “Have difficulty stopping activities or behavior when I should do so” and “Have trouble doing things in the proper order or sequence” (see Appendix D for Proposed Adult ADHD Criteria). Internal consistency of the sixitem Adult ADHD Criteria Scale used in this study was satisfactory, α = .70. World Health Organization Quality of Life Questionnaire, Brief Form (WHOQOLBREF; Bonomi, Patrick, Bushnell, & Martin, 2000) This 26-item self-report measure examines current (past two weeks) quality of life in four domains: physical health (e.g., “How well are you able to get around?”), psychological health(e.g., “To what extent do you feel your life to be meaningful?”), social relationships (e.g., “How satisfied are you with your personal relationships?”), and environmental factors (e.g., “Have you enough money to meet your needs?”). Single additional items tap general QOL (e.g., “How would you rate your quality of life?”) and general health (e.g., “How satisfied are you with your health?”) and are included in the total QOL score. Domain item responses are on a five-point scale (Not at all to Extremely); items are summed to achieve raw domain scores, with higher scores indicating better perceived QOL. Prior research suggests that internal consistency of the WHOQOLBREF domain subscales is satisfactory (Cronbach’s alpha coefficients > .7; Huang, Wu, & Frangakis, 2006). For the current sample, an abbreviated, 19-item version of the WHOQOLBREF scales was employed due to missing data in the first cohort. The social QOL scale (three items) was administered to the second cohort, only; therefore this domain is examined ADHD, Stress, and Quality of Life 23 only in that cohort and is not included in the total QOL score for the sample, as a whole. Missing data in the first cohort also included three items from the environmental QOL scale and one item from the psychological QOL scale. As a result, the psychological QOL scale was shortened to four items and the environmental scale to five items, while the physical QOL scale maintained its original seven items. Correlations between the abbreviated and full versions of the scales, using data from the second cohort, will be reported in the results section. Internal reliability for all scales on the abbreviated 19-item version of the WHOQOL-BREF, used in this study, were adequate, α = .63 (Physical), .78 (Psychological), .68 (Social), and .73 (Environmental). Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983). This frequently used measure consists of 10 items designed to assess the degree to which common life situations are deemed stressful, referencing the past month (e.g., “In the last month, how often have you been upset because of something that happened unexpectedly?”). Responses are made on a five-point scale (0 = never, 4 = very often). The PSS has previously been shown to have high internal reliability (α = .85) and to positively correlate with physical symptoms of anxiety (Cohen et al., 1983); using data from the second cohort of the current study (n = 510), internal reliability of the scale was determined to be similarly high (α = .88). Scoring of a paper-and-pencil version of the PSS, utilized for the first cohort, was such that individual item-level data is unavailable for that subsample. Brief Symptom Inventory (BSI; Derogatis, 1993). This commonly used, 53-item, selfreport measure consists of nine scales and three global indices designed to assess presence and severity of a variety of psychological symptoms (e.g., Depression, Anxiety). Examples include: “Thoughts of ending your life,” “Feelings of worthlessness” (depression symptoms), ADHD, Stress, and Quality of Life 24 “Feeling fearful,” and “Nervousness or shakiness inside” (anxiety symptoms). Symptoms are assessed by frequency or severity of symptom experience on a five-point Likert-type scale (0= Not at all, 4= Extremely). The BSI has satisfactory reliability. Internal consistency reliability for the depression and anxiety scales are .85 and .81, respectively, while test-retest reliabilities are .84 and .79 respectively (Derogatis, 1993). The predictive validity of the BSI has also been accepted as satisfactory, with a correlation of .95 between both the BSI Depression and Anxiety scales and the Symptom Checklist -90 (SCL-90) Depression and Anxiety subscales (Derogatis, 1993). Coefficient alphas were derived for the six-item Depression and Anxiety scales used in the current sample, and were similarly adequate (.87 and .78, respectively). Data Analytic Strategy Initial analyses included MANOVA comparisons of the fall and spring participant cohorts examining both independent and dependent variables to ensure statistical equivalence before combining into one sample for subsequent analyses. The primary empirical data were analyzed through a series of hierarchical multiple regressions using nine predictor variables, entered in three blocks as follows: Block One: age, gender (coded: male = 1, female = 2), and education level; Block Two: BSI Depression scale and BSI Anxiety scale; Block Three: ADHD inattentive symptoms (IA), ADHD hyperactive-impulsive symptoms (HI), ADHD Sluggish Cognitive Tempo symptoms (SCT), and Barkley and colleagues’ (2008) adult ADHD criteria (B-ADHD). The use of three predictor blocks maximizes power for variables in each block; concrete hypotheses (i.e., that ADHD symptoms will negatively relate to QOL indices and be positively associated with perceived stress) are tested in the third block. Statistically significant ADHD-related predictors in the third block, implicitly, contribute to ADHD, Stress, and Quality of Life 25 risk for negative outcomes over and above demographic makeup and other psychopathology. Six separate regressions were run to examine prediction of the following dependent variables: five quality of life indices (total score and physical, psychological, social, and environmental subdomains), and perceived stress scale score. As noted previously, the social QOL analysis included only half of the sample, resulting in lower statistical power. Other supplementary analyses (e.g., item-level regressions, effect size calculations) were employed post-hoc to further examine interesting trends and answer research questions emerging from the primary regression analyses. Results As indicated above, MANOVAs were employed to determine if differences existed between the spring and fall cohorts on both independent, Wilks’ Lambda, Λ = .98, F (8, 932) = 2.31, p = .02, and dependent variables, Λ = .43, F (5, 975) = 256.82, p < .001. Follow-up analyses indicated that the cohorts are statistically different from one another on age (an IV already included in block one of the regression model), total QOL, and perceived stress (PS), with cohort one displaying higher total QOL and cohort two higher overall age and PS. In order to exert some statistical control for potential cohort differences, a dichotomous “dummy” variable corresponding to cohort number (coded as 1 or 2) was included in the first block of indicated regression equations (i.e., for total QOL, PS), allowing for independent ascertainment of variance accounted for by other predictor variables. The versions of the psychological and environmental QOL scales used in the analyses were shortened from their original form (see description in Measures section). Correlational analyses indicated that the shortened form of both psychological QOL and environmental ADHD, Stress, and Quality of Life 26 QOL subscales are significantly and positively correlated with the original, longer version of each scale (rs = .86 and .80, respectively, p < .01, as calculated using data from cohort two). Six hierarchical multiple regressions were conducted, as planned (see above). General results from these analyses are noted in Tables 3 through 8 and are discussed further below. In the first regression predicting overall QOL, each of the three hierarchical steps added statistically significant predictive power, as noted in Table 3. The final model predicted 42% of the variance, with cohort (standardized beta, ß = -.30), depression (ß = .29), and inattention (ß = -.20) serving as strong predictors of lower levels of total QOL. Education (ß =.17) and SCT (ß = -.16) were also significant predictors. Each of the three hierarchical steps in the second regression, which examined physical QOL, also added statistically significant predictive power, as noted in Table 4. The final model predicted 15% of the variance, with SCT (ß = -.16), age (ß = -.16),depression (ß = -.15), education (ß = .10), and HI (ß = -.09) serving as significant predictors in this model. As for psychological QOL, again, each hierarchical step added meaningful predictive power (see Table 3), with the final model predicting 35% of the variance. Depression (ß = .39), IA (ß = -.20), SCT (ß = -.15), education (ß = .10), and gender (ß = -.09) were all significant predictors (see Table 5). In the fourth regression examining social QOL in the second cohort, the second and third hierarchical steps added statistically significant predictive power, as noted in Table 6. The final model predicted 5% of the variance, with depression emerging as the only significant predictor (ß = -.18). ADHD, Stress, and Quality of Life 27 In examining predictors of environmental QOL, all of the hierarchical steps added statistically significant predictive power (See Table 7). The final model predicted 22% of the variance. Depression emerged as a strong significant predictor (ß = -.24) for environmental QOL, while education and age (ß = .21 and ß = .09) were significant predictors, as well. No ADHD variable served as a statistically significant predictor for environmental QOL. As with the majority of the previous analyses, each of the three hierarchical steps in the final regression on perceived stress added statistically significant predictive power (see Table 8). The final model predicted 43% of the variance with depression (ß = .26), IA (ß = .24), cohort (ß = .15), anxiety (ß = .15),SCT (ß = .12), education (ß = -.08), gender (ß = .08), and age (ß = -.07) serving as significant predictors. Exploratory Analyses In order to further examine select, possible relationships between ADHD predictors and specific QOL scales, item-level regressions were employed post-hoc. IA and social quality of life. Hierarchical multiple regressions using two predictor variables, entered in two blocks (Block One: ADHD IA; Block Two: BSI Depression), were utilized to further determine the relationship of IA to social QOL, as well as to each item on this QOL domain scale. Depression served as the only significant predictor of social QOL in the primary analyses; thus, it was included in the exploratory analyses to determine if the effects of IA were, in effect, “overridden” by the presence of depressive symptoms. Both hierarchical steps added statistically significant predictive power. The final model predicted 4.5% of the variance, with both depression (ß = -.16) and IA (ß = -.11) serving as significant predictors of lower levels of social QOL. In an examination of associations between IA and specific items on the social QOL scale, IA significantly predicted satisfaction with personal ADHD, Stress, and Quality of Life 28 relationships (ß = -.20), sexual satisfaction (ß = -.18), and satisfaction with social support (ß = -.16), even when depression is included in the analyses (see Table 9). HI and physical quality of life. In item-level regressions examining the association of HI and each physical QOL item, HI significantly predicted energy level (β = -.29), sleep satisfaction (β = -.24), ability to perform daily activities (β = -.20), capacity for work (β = .19), and mobility (β = -.12; see Table 10). SCT and physical quality of life. An examination of associations between SCT and specific items on the physical QOL scale indicate SCT to significantly predict energy level (β = -.46), capacity for work (β = -.34), ability to perform daily activities (β = -.32), sleep satisfaction (β = -.29), and mobility (β = -.18) in this adult population (see Table 9). SCT and psychological quality of life. Exploratory analyses of the predictive powers of SCT on items in the psychological QOL scale demonstrate SCT as a significant predictor for every item on the psychological QOL subscale: ability to concentrate (β = -.46), selfsatisfaction (β = -.35), body acceptance (β = -.28), enjoyment of life (β = -.27), and meaningfulness of life (β = -.22; see Table 9). SCT and environmental quality of life. Post-hoc regression analyses were also employed on each environmental QOL item in an attempt to specify the possible relationship between SCT and environmental QOL. SCT served as a significant predictor of every aspect of environmental QOL: perceived healthiness of the physical environment (β = -.24), financial status (β = -.23), opportunity for leisure activities (β = -.22), perceived safety in daily life (β = -.19), and availability of informational resources (β = -.16; see Table 9). ADHD, Stress, and Quality of Life 29 Discussion Overall, these results indicate that the specific dimensions of ADHD symptomatology influence quality of life and perceived stress, and, importantly, that their influence is substantial even when taking potential demographic differences and depression and anxiety—two common comorbidities to ADHD—into consideration. Elaboration regarding the predictive power of various ADHD characteristics on aspects of QOL and PSS follows. Consistent with stated hypotheses, ADHD symptoms tended to be negatively associated with QOL variables and to positively correlate with perceived stress. However, while IA and SCT were fairly consistent predictors, HI symptoms predicted only one QOL variable. Inattention: Discussion of Findings IA and psychological quality of life. Inattention appears to be strongly related to lower psychological QOL. Ratings on this domain scale tap enjoyment and meaning in life, ability to concentrate, satisfaction with self and body image, as well as general mood. Ample research suggests that ADHD symptoms are related to increased prevalence of anxiety (Barkley, 2006; Waring & Lapane, 2008), depression (Bohline, 1985; Waring & Lapane, 2008), and other psychological disorders in both children and adults (see review in Barkley, 2006). More specifically, Young and Gudjonsson (2006) emphasize the importance of specifically recognizing the impact of attentional deficits on psychological functioning. Existent research already suggests that children demonstrating IA symptoms tend to be rated by teachers as being significantly more impaired in areas of psychological functioning, including anxiety, depression, withdrawal, somatic complaints, and internalizing behavior (Gaub & Carlson, 1997). Inattention has been more broadly linked to academic (Barkley, ADHD, Stress, and Quality of Life 3