- Email: [email protected]
Depression, Anxiety, and Quality of Life in Patients With Atrial Fibrillation
Original Research ATRIAL FIBRILLATION
Depression, Anxiety, and Quality of Life in Patients With Atrial Fibrillation* Graham Thrall, PhD; Gregory Y. H. Lip, MD; Douglas Carroll, PhD; and Deirdre Lane, PhD
Objective: To examine the prevalence and persistence of depression and anxiety in patients with atrial fibrillation (AF), and their effect on future quality of life (QoL) status. Methods: The Beck Depression Inventory and State-Trait Anxiety Inventory were completed by 101 patients with AF (62 men; mean age ⴞ SD, 66.3 ⴞ 11.0 years), who were compared to 97 patients with hypertension (as “disease control” subjects) in sinus rhythm (64 men; mean age, 68.0 ⴞ 7.2 years) at baseline and at 6 months. QoL was ascertained at both time points using Dartmouth Care Cooperative Information Project charts. Results: At baseline among AF patients, symptoms of depression, state anxiety, and trait anxiety prevailed in 38%, 28%, and 38%, respectively; analogous data for hypertensive patients were 30%, 23%, and 22%. AF patients displayed higher levels of trait anxiety (p < 0.05), with no significant differences in baseline depression, state anxiety, and QoL between patients with AF and disease control subjects. Symptoms of depression and anxiety (state and trait) persisted at 6 months in 36.8% and 33.3%, respectively. Symptoms of depression (p < 0.001) and anxiety (p < 0.001) at baseline, female gender (p ⴝ 0.01), ethnicity (p ⴝ 0.01), and employment status (p ⴝ 0.03) were significantly correlated with QoL at 6 months in the patients with AF. Multiple regression analysis revealed that baseline depression score provided the best independent prediction of 6-month QoL (R2 ⴝ 0.20), although gender and employment status also entered the model. Conclusion: Approximately one third of AF patients have elevated levels of depression and anxiety, which persist at 6 months. Symptoms of depression were the strongest independent predictor of future QoL in these patients. (CHEST 2007; 132:1259 –1264) Key words: anxiety; atrial fibrillation; depression; hypertension; quality of life Abbreviations: AF ⫽ atrial fibrillation; BDI ⫽ Beck Depression Inventory; CI ⫽ confidence interval; COOP ⫽ Care Cooperative Information Project; SF-36 ⫽ Short Form-36; STAI ⫽ Stait-Trait Anxiety Inventory; OR ⫽ odds ratio; QoL ⫽ quality of life
the epidemiology, cost, clinical conseA lthough quences, and efficacy of various treatment regimes for atrial fibrillation (AF) have been subject to considerable study,1–3 less attention has been paid to *From the University Department of Medicine, City Hospital, Birmingham, UK. The authors have no conflicts of interest to declare. Manuscript received January 10, 2007; revision accepted June 2, 2007. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Gregory Y. H. Lip, MD, University Department of Medicine, City Hospital, Birmingham, B18 7QH UK; e-mail: [email protected] DOI: 10.1378/chest.07-0036 www.chestjournal.org
patient-related issues, especially psychological morbidity and quality of life (QoL). To date, no study has specifically examined the levels of depression and anxiety in patients with AF. However, patients with AF have significantly poorer QoL compared to healthy control subjects,4 – 6 the general population,7 and other coronary heart disease patients.4 Our systematic review of the literature8 demonstrated that many studies assessing QoL were compromised by methodologic weaknesses, including small sample size, not employing a control group, and use of nonvalidated tools to assess QoL. Also, many were performed on subgroups of clinical trial patients, resulting in biases, and may not have been adequately powered for such analyses.8 CHEST / 132 / 4 / OCTOBER, 2007
1259
We hypothesized that AF patients would report poorer QoL than a hypertensive “disease control” group in sinus rhythm. We further hypothesized that QoL in both patient groups would be predicted by depression and anxiety, and that these associations would be reliable across time and independent of demographic status. Finally, we examined the change in QoL across the 6 months of the study, as analogous changes in depression and anxiety would enable us to test the intimacy of the association between the affective disposition and QoL. Materials and Methods All consecutive patients with a diagnosis of AF attending our specialist AF cardiology clinic were eligible for inclusion (Table 1). Patients were excluded for the following: (1) age ⬍ 18 years; (2) previous nonpharmacologic intervention (excluding direct current cardioversion) to correct this arrhythmia; (3) malignancy of any type; (4) myocardial infarction, transient ischemic attack/ stroke, coronary artery bypass graft surgery, or percutaneous transluminal coronary angioplasty within the previous 6 months; and (5) not able to read English. AF patients were age matched
and sex matched with a disease control group in sinus rhythm (patients with essential hypertension [BP ⬎ 160/90 mm Hg untreated, or established on antihypertensive drugs]). Hypertensive patients in sinus rhythm were chosen as the disease control patients because hypertension is one of the most common comorbidities associated with AF, and we hope to determine whether AF per se is related to depression and anxiety. During the study period, 195 AF and 160 hypertensive patients were approached to participate. Of these, 95 AF patients (48.7%) and 63 hypertensive patients (39.4%) refused to participate. Consequently, 101 of the AF patients and 97 of the hypertensive patients provided written informed consent, completed the baseline questionnaires, and constitute the effective study population. There were few differences between the AF and hypertensive patients who participated and those who declined, although nonparticipating AF patients (17.2%) and hypertensive patients (21.8%) were significantly more likely to be nonwhite (p ⫽ 0.02 and p ⬍ 0.01, respectively). No other significant differences in demographic and clinical characteristics between participating and nonparticipating AF patients were observed (data not shown). Procedure Social, demographic, and clinical details were recorded. Townsend deprivation scores were used as a measure of socio-
Table 1—Demographic and Clinical Characteristics of the AF and Hypertensive Patients* Characteristics Baseline demographic characteristics Mean (SD), age yr Male gender Ethnicity White Afro-Caribbean South-Asian Occupational status Mean (SD) deprivation score Baseline clinical characteristics Significant comorbidity One Two or more Hypertension Diabetes mellitus Myocardial infarction Stroke Transient ischemic attack Current medication Warfarin Antiplatelet Diuretic Calcium channel blocker -Blocker ␣-Blocker Angiotensin-converting enzyme inhibitor Angiotensin II receptor blocker Statin
AF Patients (n ⫽ 101)
Hypertensive Patients (n ⫽ 97)
p Value
66.3 (11.0) 62 (61.4)
68.0 (7.2) 64 (66.0)
0.23 0.50
96 (95.0) 3 (3.0) 2 (2.0) 38 (37.6) 4.37 (3.8)
87 (89.7) 9 (9.3) 1 (1.0) 29.9 (29.9) 4.74 (3.6)
57 (56.4) 39 (38.6) 89 (88.1) 11 (10.9) 6 (5.9) 6 (5.9) 6 (5.9)
50 (51.5) 20 (20.6) 97 (100) 47 (48.5) 11 (11.3) 5 (5.2) 6 (6.2)
0.15 ⬍0.01 ⬍0.01 0.27 1.00 1.00
75 (74.3) 22 (22.8) 41 (40.6) 49 (48.5) 56 (55.4) 7 (6.9) 32 (31.7) 24 (23.8) 30 (29.7)
2 (2.1) 42 (43.3) 51 (52.6) 39 (40.2) 49 (50.5) 53 (54.6) 42 (43.3) 13 (13.4) 57 (58.8)
⬍0.001 0.001 0.16 0.24 0.68 ⬍0.001 0.12 0.05 ⬍0.001
0.24 0.40 0.58
*Data are presented as No. (%) unless otherwise indicated. The presence of AF was defined as the absence of a P-wave in association with rapid oscillations or fibrillatory waves on an ECG. Recurrent AF patients included patients with paroxysmal and persistent AF. Paroxysmal AF was defined as patients with documented paroxysms of AF lasting ⱖ 10 beats on 24-h Holter monitoring, while persistent AF was defined as patients where cardioversion to sinus rhythm was being considered. Permanent AF was defined as AF being present for ⱖ 1 yr and cardioversion was considered inappropriate or was previously unsuccessful. 1260
Original Research
economic status, based on an individual’s postal code.9 Patients completed a battery of questionnaires on two separate occasions: baseline and 6 months. The questionnaire pack comprised the following: the Beck Depression Inventory (BDI),10 the Stait-Trait Anxiety Inventory (STAI),11 and the Dartmouth Care Cooperative Information Project (COOP) chart system.12 Scores on the BDI, STAI, and Dartmouth (COOP) chart system ranged from 0 to 63, 20 to 80, and 8 to 40, respectively, with higher scores indicating greater levels of depressive symptoms, higher levels of anxiety, and a poorer QoL. BDI scores ⱖ 10 and trait anxiety scores ⱖ 40 were considered to indicate significant symptoms of depression and anxiety, criteria that has been employed previously in studies13–15 examining depression and anxiety in patients with coronary heart disease. The BDI and STAI have been used extensively to examine symptoms of depression and anxiety in patient groups and the Dartmouth COOP charts are easy to understand and complete and provide an overall QoL score. All questionnaires employed had acceptable psychometric properties. Data Analysis All statistical tests were two tailed; p values ⱕ 0.05 were considered statistically significant. Dichotomous variables were coded as follows: patient type (AF ⫽ 1, hypertension ⫽ 2), gender (female ⫽ 0, male ⫽ 1), and employment (employed ⫽ 0, unemployed ⫽ 1). Analysis was by linear regression, with a hierarchical approach applied in multivariate analysis. The following cross-sectional analytic strategy was applied at both baseline and follow-up. First, a simple bivariate model was tested to determine whether either depression or trait anxiety were associated with QoL. Second, a model was tested in which depression was entered at step one followed by patient type and its interaction with depression at step two. In calculating this interaction term, the BDI values were mean centered to avoid multicollinearity.16 The patient type variable allowed us to test whether the patient groups differed with regard to QoL, and the interaction term tested whether the association between the affective disposition and QoL varied between the two patient groups. At step three, the following possible demographic confounders were entered: age, gender, deprivation score, and employment status. Finally, 6-month change scores were calculated for depression, trait anxiety, and QoL to test whether changes in QoL reflected changes in affective status. The same regression strategy was applied using these change scores.
Results The majority of AF subjects were male (61.4%), white (89.7%), and unemployed (70.1%), with a median duration of diagnosed AF being 21 months (interquartile range, 8 to 49 months). Participants had recurrent AF in 58.4% (paroxysmal AF, 52.5%; persistent AF, 5.9%) and permanent AF in 41.6%. Hypertension was the most common coexisting cardiovascular comorbidity (88.1%), followed by diabetes mellitus (10.9%). The majority of AF patients were receiving some form of anticoagulant (warfarin, 72.3%) or antiplatelet therapy (22.8%). No differences were observed in age, gender, ethnicity, occupational status, and deprivation score. Hypertensive patients were more likely to have diabetes mellitus (p ⬍ 0.01), but other cardiovascular comorbidities were not significantly different (Table 1). www.chestjournal.org
Depression, Anxiety, and QoL at Baseline AF and hypertensive patients displayed similar levels of depression (mean score, and number of scores ⱖ 10 on BDI) and QoL. The relationship between depression and diabetes mellitus was examined, given the significant difference in the prevalence of diabetes mellitus between AF and hypertensive patients. Presence of diabetes mellitus did not predict baseline depression in either group (odds ratio [OR], 0.28; 95% confidence interval [CI], ⫺ 4.26 to 4.83 [p ⫽ 0.902]; and OR, ⫺ 0.87; 95% CI, ⫺ 3.43 to 1.68 [p ⫽ 0.498] for AF and hypertensive patients, respectively) or for both groups combined (OR, ⫺ 1.27; 95% CI, ⫺ 3.36 to 0.82; p ⫽ 0.23). Higher levels of trait anxiety (mean score, p ⬍ 0.02) and percentage of scores ⱖ 40 on STAI, p ⫽ 0.03) were observed in the AF patients (Table 2). Depression and trait anxiety were highly comorbid conditions in AF patients, with 71% of patients reporting BDI scores ⱖ 10 also exhibiting high levels of anxiety. Female AF patients had a poorer QoL than male AF patients (t[99] ⫽ 3.43; p ⬍ 0.05), but no significant differences in any of the psychological parameters were observed between recurrent and permanent AF patients (data not shown). AF and hypertensive patients displaying significant symptoms of depression were demographically and clinically comparable to nondepressed patients
Table 2—Baseline and 6-Month Psychological Characteristics of the AF and Hypertensive Patients* Variables Baseline psychological characteristics Median (IQR) BDI score BDI score ⱖ 10 Mean (SD) trait anxiety score Trait anxiety score ⱖ 40 Mean (SD) QoL score 6-month psychological characteristics Median (IQR) BDI score BDI score ⱖ 10 Mean (SD) trait anxiety score Trait anxiety score ⱖ 40 Mean (SD) QoL score
AF Patients
Hypertensive Patients
p Value
7.0 (3–13)
6.0 (2–10)
0.07
38 (37.6) 37.4 (12.6)
29 (30.0) 33.3 (11.4)
0.32 0.02
38 (37.6)
21 (21.6)
0.03
20.4 (5.7)
19.9 (5.3)
0.57
7.0 (3–13)
6.0 (2–11)
0.61
28 (36.8) 36.9 (12.5)
20 (29.4) 32.6 (11.2)
0.48 0.03
24 (33.3)
17 (25.4)
0.25
20.3 (6.2)
20.5 (6.2)
0.95
*Data are presented as No. (%) unless otherwise indicated. IQR ⫽ interquartile range. CHEST / 132 / 4 / OCTOBER, 2007
1261
(p ⫽ not significant; data not shown). AF and hypertensive patients displaying elevated symptoms of trait anxiety were found to be younger (p ⫽ 0.02 and p ⬍ 0.01, respectively), with anxious hypertensive patients also more likely to be female (p ⬍ 0.01, data not shown). Predictors of QoL at Baseline Bivariate analysis indicated a strong association between depression and baseline QoL ( ⫽ 0.58, t ⫽ 9.78, p ⬍ 0.001); the higher the depression score the poorer the QoL. This association survived in the model also entering patient type and the interaction between depression and interaction score ( ⫽ 0.49), as well as adjustment for demographic factors ( ⫽ 0.50). Neither patient type ( ⫽ 0.03) nor the interaction ( ⫽ 0.10) were associated with QoL. Of the demographic variables, men and those in paid employment report better QoL than women ( ⫽ ⫺ 0.23) and those who were unemployed ( ⫽ 0.17). Trait anxiety was similarly associated with QoL ( ⫽ 0.56, t ⫽ 9.28, p ⬍ 0.001), and although the association was attenuated, it survived entry for patient type and the interaction term ( ⫽ 0.42) and demographic status ( ⫽ 0.46). Again, there were no differences between patient group in QoL ( ⫽ 0.04), nor was there an interaction between trait anxiety and patient type ( ⫽ 0.15). In a final one-step model in which depression and trait anxiety were entered, both emerged as significant predictors of contemporary QoL ( ⫽ 0.38, t ⫽ 3.99, p ⬍ 0.001; and  ⫽ 0.26, t ⫽ 2.77, p ⫽ 0.006, respectively). Persistence of Depression and Anxiety at 6 Months, and Stability of QoL The BDI and STAI were returned and completed at 6 months by 76 AF patients (75.2%) and 72 AF patients (71.3%), respectively. At 6 months, depression was present in 28 patients (36.8%) [Table 2], with 52.6% of patients reporting significant depression at baseline continuing to report elevated depressive symptoms at follow-up. Elevated levels of trait anxiety at follow-up were present in 24 patients (33.3%). Of the AF patients reporting high levels of anxiety at baseline, 52.6% continued to report elevated trait anxiety at 6 months. QoL at 6 months was strongly correlated with QoL at baseline (r ⫽ 0.67, p ⬍ 0.001), with the same association holding true for depression (r ⫽ 0.79, p ⬍ 0.001) and trait anxiety (r ⫽ 0.81, p ⬍ 0.001). QoL, depression, and anxiety did not change significantly between baseline and follow-up. The significant baseline difference in mean trait anxiety score 1262
between AF and hypertensive persisted at 6 months, although the difference in the dichotomized variable was no longer significant (p ⫽ 0.25). Predictors of QoL at 6 Months Outcomes were similar to those found at baseline. Depression at follow-up was strongly associated with contemporary QoL ( ⫽ 0.59, t ⫽ 8.52, p ⬍ 0.001). This association was not attenuated in a model adjusting for patient type and its interaction with depression ( ⫽ 0.62), and in a model additionally adjusting for demographic status ( ⫽ 0.65). Again, neither patient type ( ⫽ 0.04) nor the interaction ( ⫽ ⫺ 0.04) were associated with QoL. Of the demographic variables examined, gender ( ⫽ ⫺ 0.13) and employment status ( ⫽ 0.20) were related to QoL. Trait anxiety at 6 months was similarly related to 6-month QoL ( ⫽ 48, t ⫽ 6.34, p ⬍ 0.001). Again, the regression coefficients remained high in subsequent models adjusting for patient status and demographics ( ⫽ 0.39 in both cases). In contrast to the baseline findings, only depression emerged as a predictor of QoL in the model that entered both depression ( ⫽ 0.60, t ⫽ 4.94, p ⬍ 0.001) and trait anxiety ( ⫽ 0.00). Predictors of the Change in QoL Over 6-Month Follow-up Change in QoL between baseline and 6 months was significantly related to change in depression score ( ⫽ 0.37, t ⫽ 4.68, p ⬍ 0.001); improvements in depression were associated with enhanced QoL. This association was not influenced by patient type ( ⫽ 0.36) nor by baseline demographic characteristics ( ⫽ 0.40). Neither patient type nor demographic status was related to the direction and extent of change in QoL. The same was true for change in trait anxiety ( ⫽ 0.35, t ⫽ 4.22, p ⬍ 0.001); in this case, the association was abolished in models that included patient type and demographic status.
Discussion Depression and anxiety have been found to be strong predictor of QoL in myocardial infarction patients.17 Although QoL has been extensively studied in patients with AF,8 depression and anxiety have yet to receive attention. There has been no previous examination of the association of the affective status of AF patients and their QoL. Rates of depression and trait anxiety observed in our AF cohort were comparable to that seen in patients following myocardial infarction,13,18 –22 imOriginal Research
plying similar levels of psychological morbidity characterize these two patient groups. In the present study, hypertensive patients presented with similar levels of depression to patients with AF, although they exhibited lower levels of trait anxiety. In contrast to AF patients, the prevalence of depression in hypertensive patients has received some attention.23–25 Rabkin et al24 demonstrated that major depression was three times more common among hypertensive patients even when known cardiac risk factors were taken into account. Mild depressive symptoms are present in 20% of hypertensive patients.25 Significant symptoms of depression and trait anxiety persisted over the first 6 months of follow-up (Table 2), with approximately one half of all patients reporting elevated symptoms of depression and anxiety at baseline, also reporting symptoms similarly increased at 6 months. In addition, the average depression and trait anxiety scores showed no attenuation from baseline to follow-up. Such observations suggest that AF patients experience significant psychological morbidity, which is protracted and not limited to a single point of observation. The current study reveals no significant baseline differences in QoL between patients with AF and hypertensive patients in sinus rhythm. We are only aware of one previous study25 that examined QoL in AF patients and a hypertensive control group, which reported a lower QoL on six of the eight scales of the Short Form-36 (SF-36) in AF patients compared to hypertensive control subjects. It is plausible that this apparent discrepancy in results arises from the exaggerated baseline scores in the previous study,26 due to the highly symptomatic patients undergoing an invasive procedure (pulmonary vein isolation). Differences in QoL between patients with recurrent AF and permanent AF have not been subject to investigation. The current study, although not intentionally powered to detect differences between the subgroups of AF patients, demonstrated that patients with paroxysmal AF showed comparable levels of QoL to patients with permanent AF. Only one of the five observational studies4 –7,27 previously examining QoL did so exclusively in a cohort of permanent AF patients, and found that men with permanent AF displayed similar levels of QoL (SF-36) to an aged-matched control group in sinus rhythm.5 The authors argued that having a predictable clinical course of rate control and anticoagulation, in addition to relatively well-controlled heart rates, might explain the equivocal findings. However, one study28 demonstrated that there was no significant association between achieved heart rate either at rest or during exercise and QoL. Conversely, the significant impairment in QoL reported in paroxysmal AF www.chestjournal.org
patients4 – 6,27 may be a consequence of rapid heart rates that are potentially more symptom producing than permanent (rate-controlled) AF, or individuals perceiving their illness to be more intrusive on their everyday lives.29 Female AF patients had a poorer QoL than men. Previous research on gender differences in QoL in AF patients demonstrated that female patients display poorer physical component summary scores on the SF-36 than their male counterparts.27 Although this result was confounded by the fact that female patients were significantly older than male patients, such discrepancies may not fully account for gender differences in QoL.27 Indeed, the greater impact of AF on QoL in women may be attributed to their heightened sensitivity to the disease and its associated symptoms, gender differences in the perception of illness, or a lower threshold for reporting illness burden.27 Increased sensitivity to symptoms is unlikely to be a major factor driving these differences because the impairment of QoL experienced by women persisted even after controlling for somatization scores.27 Depression and anxiety were strongly associated with QoL in our cross-sectional analysis. However, our QoL measure did not include the affective item in the Dartmouth COOP chart system. Further, the associations emerged at both baseline and follow-up, were not influenced by patient type, and survived adjustment for a number of sociodemographic variables. Whereas at baseline, depression and anxiety appeared to be independently related to QoL, at 6 months only depression emerged from analysis entering both independent variables. This resonates with the results from a study13,17 of QoL in myocardial infarction patients, in which QoL is more consistently predicted by depression than by anxiety. The intimacy of the association between depression and QoL is best demonstrated in the present study by the change score analysis. Changes in depression over 6 months were significantly associated with changes in QoL, even taking into account possible confounding variables; this was not the case for anxiety. This holds strong implications for the psychological management of AF patients. Our results suggest that depression should be the focus because reductions in depression should translate into improvements in QoL. Increasingly, QoL is regarded as an important clinical outcome in the management of chronic conditions, and the present study suggests a strategy for improving QoL in AF patients may be needed. In conclusion, approximately one third of AF patients have elevated levels of depression and anxiety, which persist at 6 months. Symptoms of depresCHEST / 132 / 4 / OCTOBER, 2007
1263
sion were the strongest independent predictor of future QoL in these patients. References 1 Bilal IM, Taneja AK, Lip GY, et al. Recent developments in atrial fibrillation. BMJ 2005; 330:238 –243 2 Dewilde S, Carey IM, Emmas C, et al. Trends in the prevalence of diagnosed atrial fibrillation, its treatment with anticoagulation, and predictors of such treatment in UK primary care. Heart 2006; 92:1064 –1070 3 Ruigomez A, Johansson S, Wallander MA, et al. Incidence of chronic atrial fibrillation in general practice and its treatment patterns. J Clin Epidemiol 2002; 55:358 –363 4 Dorian P, Jung W, Newman D, et al. The impairment of health-related quality of life in patients with intermittent atrial fibrillation: implications for assessment of investigational therapy. J Am Coll Cardiol 2000; 36:1303–1309 5 Howes CJ, Reid MC, Brandt C, et al. Exercise tolerance and quality of life in elderly patients with chronic atrial fibrillation. J Cardiovasc Pharmacol Ther 2001; 6:23–29 6 van den Berg MP, Hassink RJ, Tuinenburg AE, et al. Quality of life in patients with paroxysmal atrial fibrillation and its predictors: importance of the autonomic system. Eur Heart J 2001; 22:247–253 7 Kang Y, Bahler R. Health related quality of life in patients newly diagnosed with atrial fibrillation. Eur J Cardiovasc Nurs 2003; 3:71–76 8 Thrall G, Lane D, Carroll D, et al. Quality of life in patients with atrial fibrillation: a systematic review. Am J Med 2006; 119:448.e1– e19 9 Townsend P. Deprivation. J Soc Policy 1987; 16:125–146 10 Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561–571 11 Spielberger CD, Gorsuch RL, Lushene R, et al. State-Trait Anxiety Inventory for adults. Palo Alto, CA: Consulting Psychologists Press, 1983 12 Nelson E, Wasson J, Kirk J, et al. Assessment of function in routine clinical practice: description of the COOP chart method and preliminary findings. J Chron Dis 1987; 40:55S– 63S 13 Lane D, Carroll D, Ring C, et al. Mortality and quality of life 12 months after myocardial infarction: effects of depression and anxiety. Psychosom Med 2001; 63:221–230 14 Frasure-Smith N, Lesperance F, Gravel G, et al. Depression and health-care costs during the first year following myocardial infarction. J Psychosom Res 2000; 48:471– 478
1264
15 Frasure-Smith N, Lesperance F, Talajic M. Depression and 18-month prognosis after myocardial infarction. Circulation 1995; 91:999 –1005 16 Akin LS, West SG. Multiple regression: testing and interpreting interactions. London, UK: Sage, 1991 17 Lane D, Carroll D, Ring C, et al. Effects of depression and anxiety on mortality and quality-of-life 4 months after myocardial infarction. J Psychosom Res 2000; 49:229 –238 18 Frasure-Smith N, Lesperance F, Talajic M. The impact of negative emotions on prognosis following myocardial infarction: is it more than depression? Health Psychol 1995; 14:388 –398 19 Frasure-Smith N, Lesperance F, Juneau M, et al. Gender, depression, and one-year prognosis after myocardial infarction. Psychosom Med 1999; 61:26 –37 20 Lane D, Carroll D, Ring C, et al. In-hospital symptoms of depression do not predict mortality 3 years after myocardial infarction. Int J Epidemiol 2002; 31:1179 –1182 21 Ahern DK, Gorkin L, Anderson JL, et al. Biobehavioral variables and mortality or cardiac arrest in the Cardiac Arrhythmia Pilot Study (CAPS). Am J Cardiol 1990; 66: 59 – 62 22 Denollet J, Brutsaert DL. Personality, disease severity, and the risk of long-term cardiac events in patients with a decreased ejection fraction after myocardial infarction. Circulation 1998; 97:167–173 23 Scalco AZ, Scalco MZ, Azul JB, et al. Hypertension and depression. Clinics 2005; 60:241–250 24 Rabkin JG, Charles E, Kass F. Hypertension and DSM-III depression in psychiatric outpatients. Am J Psychiatry 1983; 140:1072–1074 25 Bonnet F, Irving K, Terra JL, et al. Depressive symptoms are associated with unhealthy lifestyles in hypertensive patients with the metabolic syndrome. J Hypertens 2005; 23:611– 617 26 Nilsson B, Chen X, Svendsen JH. Effects of pulmonary vein isolation on quality of life in patients with paroxysmal atrial fibrillation. Heart Drug 2003; 3:173–179 27 Paquette M, Roy D, Talajic M, et al. Role of gender and personality on quality-of-life impairment in intermittent atrial fibrillation. Am J Cardiol 2000; 86:764 –768 28 Cooper HA, Bloomfield DA, Bush DE, et al. Relationship between achieved heart rate and outcomes in patients with atrial fibrillation (AFFIRM Study). Am J Cardiol 2004; 93:1247–1253 29 Devins GM. Illness intrusiveness and the psychosocial impact of lifestyle disruptions in chronic life-threatening disease. Adv Ren Replace Ther 1994; 1:251–263
Original Research
Depression, Anxiety, and Quality of Life in Patients With Atrial Fibrillation* Graham Thrall, PhD; Gregory Y. H. Lip, MD; Douglas Carroll, PhD; and Deirdre Lane, PhD
Objective: To examine the prevalence and persistence of depression and anxiety in patients with atrial fibrillation (AF), and their effect on future quality of life (QoL) status. Methods: The Beck Depression Inventory and State-Trait Anxiety Inventory were completed by 101 patients with AF (62 men; mean age ⴞ SD, 66.3 ⴞ 11.0 years), who were compared to 97 patients with hypertension (as “disease control” subjects) in sinus rhythm (64 men; mean age, 68.0 ⴞ 7.2 years) at baseline and at 6 months. QoL was ascertained at both time points using Dartmouth Care Cooperative Information Project charts. Results: At baseline among AF patients, symptoms of depression, state anxiety, and trait anxiety prevailed in 38%, 28%, and 38%, respectively; analogous data for hypertensive patients were 30%, 23%, and 22%. AF patients displayed higher levels of trait anxiety (p < 0.05), with no significant differences in baseline depression, state anxiety, and QoL between patients with AF and disease control subjects. Symptoms of depression and anxiety (state and trait) persisted at 6 months in 36.8% and 33.3%, respectively. Symptoms of depression (p < 0.001) and anxiety (p < 0.001) at baseline, female gender (p ⴝ 0.01), ethnicity (p ⴝ 0.01), and employment status (p ⴝ 0.03) were significantly correlated with QoL at 6 months in the patients with AF. Multiple regression analysis revealed that baseline depression score provided the best independent prediction of 6-month QoL (R2 ⴝ 0.20), although gender and employment status also entered the model. Conclusion: Approximately one third of AF patients have elevated levels of depression and anxiety, which persist at 6 months. Symptoms of depression were the strongest independent predictor of future QoL in these patients. (CHEST 2007; 132:1259 –1264) Key words: anxiety; atrial fibrillation; depression; hypertension; quality of life Abbreviations: AF ⫽ atrial fibrillation; BDI ⫽ Beck Depression Inventory; CI ⫽ confidence interval; COOP ⫽ Care Cooperative Information Project; SF-36 ⫽ Short Form-36; STAI ⫽ Stait-Trait Anxiety Inventory; OR ⫽ odds ratio; QoL ⫽ quality of life
the epidemiology, cost, clinical conseA lthough quences, and efficacy of various treatment regimes for atrial fibrillation (AF) have been subject to considerable study,1–3 less attention has been paid to *From the University Department of Medicine, City Hospital, Birmingham, UK. The authors have no conflicts of interest to declare. Manuscript received January 10, 2007; revision accepted June 2, 2007. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Gregory Y. H. Lip, MD, University Department of Medicine, City Hospital, Birmingham, B18 7QH UK; e-mail: [email protected] DOI: 10.1378/chest.07-0036 www.chestjournal.org
patient-related issues, especially psychological morbidity and quality of life (QoL). To date, no study has specifically examined the levels of depression and anxiety in patients with AF. However, patients with AF have significantly poorer QoL compared to healthy control subjects,4 – 6 the general population,7 and other coronary heart disease patients.4 Our systematic review of the literature8 demonstrated that many studies assessing QoL were compromised by methodologic weaknesses, including small sample size, not employing a control group, and use of nonvalidated tools to assess QoL. Also, many were performed on subgroups of clinical trial patients, resulting in biases, and may not have been adequately powered for such analyses.8 CHEST / 132 / 4 / OCTOBER, 2007
1259
We hypothesized that AF patients would report poorer QoL than a hypertensive “disease control” group in sinus rhythm. We further hypothesized that QoL in both patient groups would be predicted by depression and anxiety, and that these associations would be reliable across time and independent of demographic status. Finally, we examined the change in QoL across the 6 months of the study, as analogous changes in depression and anxiety would enable us to test the intimacy of the association between the affective disposition and QoL. Materials and Methods All consecutive patients with a diagnosis of AF attending our specialist AF cardiology clinic were eligible for inclusion (Table 1). Patients were excluded for the following: (1) age ⬍ 18 years; (2) previous nonpharmacologic intervention (excluding direct current cardioversion) to correct this arrhythmia; (3) malignancy of any type; (4) myocardial infarction, transient ischemic attack/ stroke, coronary artery bypass graft surgery, or percutaneous transluminal coronary angioplasty within the previous 6 months; and (5) not able to read English. AF patients were age matched
and sex matched with a disease control group in sinus rhythm (patients with essential hypertension [BP ⬎ 160/90 mm Hg untreated, or established on antihypertensive drugs]). Hypertensive patients in sinus rhythm were chosen as the disease control patients because hypertension is one of the most common comorbidities associated with AF, and we hope to determine whether AF per se is related to depression and anxiety. During the study period, 195 AF and 160 hypertensive patients were approached to participate. Of these, 95 AF patients (48.7%) and 63 hypertensive patients (39.4%) refused to participate. Consequently, 101 of the AF patients and 97 of the hypertensive patients provided written informed consent, completed the baseline questionnaires, and constitute the effective study population. There were few differences between the AF and hypertensive patients who participated and those who declined, although nonparticipating AF patients (17.2%) and hypertensive patients (21.8%) were significantly more likely to be nonwhite (p ⫽ 0.02 and p ⬍ 0.01, respectively). No other significant differences in demographic and clinical characteristics between participating and nonparticipating AF patients were observed (data not shown). Procedure Social, demographic, and clinical details were recorded. Townsend deprivation scores were used as a measure of socio-
Table 1—Demographic and Clinical Characteristics of the AF and Hypertensive Patients* Characteristics Baseline demographic characteristics Mean (SD), age yr Male gender Ethnicity White Afro-Caribbean South-Asian Occupational status Mean (SD) deprivation score Baseline clinical characteristics Significant comorbidity One Two or more Hypertension Diabetes mellitus Myocardial infarction Stroke Transient ischemic attack Current medication Warfarin Antiplatelet Diuretic Calcium channel blocker -Blocker ␣-Blocker Angiotensin-converting enzyme inhibitor Angiotensin II receptor blocker Statin
AF Patients (n ⫽ 101)
Hypertensive Patients (n ⫽ 97)
p Value
66.3 (11.0) 62 (61.4)
68.0 (7.2) 64 (66.0)
0.23 0.50
96 (95.0) 3 (3.0) 2 (2.0) 38 (37.6) 4.37 (3.8)
87 (89.7) 9 (9.3) 1 (1.0) 29.9 (29.9) 4.74 (3.6)
57 (56.4) 39 (38.6) 89 (88.1) 11 (10.9) 6 (5.9) 6 (5.9) 6 (5.9)
50 (51.5) 20 (20.6) 97 (100) 47 (48.5) 11 (11.3) 5 (5.2) 6 (6.2)
0.15 ⬍0.01 ⬍0.01 0.27 1.00 1.00
75 (74.3) 22 (22.8) 41 (40.6) 49 (48.5) 56 (55.4) 7 (6.9) 32 (31.7) 24 (23.8) 30 (29.7)
2 (2.1) 42 (43.3) 51 (52.6) 39 (40.2) 49 (50.5) 53 (54.6) 42 (43.3) 13 (13.4) 57 (58.8)
⬍0.001 0.001 0.16 0.24 0.68 ⬍0.001 0.12 0.05 ⬍0.001
0.24 0.40 0.58
*Data are presented as No. (%) unless otherwise indicated. The presence of AF was defined as the absence of a P-wave in association with rapid oscillations or fibrillatory waves on an ECG. Recurrent AF patients included patients with paroxysmal and persistent AF. Paroxysmal AF was defined as patients with documented paroxysms of AF lasting ⱖ 10 beats on 24-h Holter monitoring, while persistent AF was defined as patients where cardioversion to sinus rhythm was being considered. Permanent AF was defined as AF being present for ⱖ 1 yr and cardioversion was considered inappropriate or was previously unsuccessful. 1260
Original Research
economic status, based on an individual’s postal code.9 Patients completed a battery of questionnaires on two separate occasions: baseline and 6 months. The questionnaire pack comprised the following: the Beck Depression Inventory (BDI),10 the Stait-Trait Anxiety Inventory (STAI),11 and the Dartmouth Care Cooperative Information Project (COOP) chart system.12 Scores on the BDI, STAI, and Dartmouth (COOP) chart system ranged from 0 to 63, 20 to 80, and 8 to 40, respectively, with higher scores indicating greater levels of depressive symptoms, higher levels of anxiety, and a poorer QoL. BDI scores ⱖ 10 and trait anxiety scores ⱖ 40 were considered to indicate significant symptoms of depression and anxiety, criteria that has been employed previously in studies13–15 examining depression and anxiety in patients with coronary heart disease. The BDI and STAI have been used extensively to examine symptoms of depression and anxiety in patient groups and the Dartmouth COOP charts are easy to understand and complete and provide an overall QoL score. All questionnaires employed had acceptable psychometric properties. Data Analysis All statistical tests were two tailed; p values ⱕ 0.05 were considered statistically significant. Dichotomous variables were coded as follows: patient type (AF ⫽ 1, hypertension ⫽ 2), gender (female ⫽ 0, male ⫽ 1), and employment (employed ⫽ 0, unemployed ⫽ 1). Analysis was by linear regression, with a hierarchical approach applied in multivariate analysis. The following cross-sectional analytic strategy was applied at both baseline and follow-up. First, a simple bivariate model was tested to determine whether either depression or trait anxiety were associated with QoL. Second, a model was tested in which depression was entered at step one followed by patient type and its interaction with depression at step two. In calculating this interaction term, the BDI values were mean centered to avoid multicollinearity.16 The patient type variable allowed us to test whether the patient groups differed with regard to QoL, and the interaction term tested whether the association between the affective disposition and QoL varied between the two patient groups. At step three, the following possible demographic confounders were entered: age, gender, deprivation score, and employment status. Finally, 6-month change scores were calculated for depression, trait anxiety, and QoL to test whether changes in QoL reflected changes in affective status. The same regression strategy was applied using these change scores.
Results The majority of AF subjects were male (61.4%), white (89.7%), and unemployed (70.1%), with a median duration of diagnosed AF being 21 months (interquartile range, 8 to 49 months). Participants had recurrent AF in 58.4% (paroxysmal AF, 52.5%; persistent AF, 5.9%) and permanent AF in 41.6%. Hypertension was the most common coexisting cardiovascular comorbidity (88.1%), followed by diabetes mellitus (10.9%). The majority of AF patients were receiving some form of anticoagulant (warfarin, 72.3%) or antiplatelet therapy (22.8%). No differences were observed in age, gender, ethnicity, occupational status, and deprivation score. Hypertensive patients were more likely to have diabetes mellitus (p ⬍ 0.01), but other cardiovascular comorbidities were not significantly different (Table 1). www.chestjournal.org
Depression, Anxiety, and QoL at Baseline AF and hypertensive patients displayed similar levels of depression (mean score, and number of scores ⱖ 10 on BDI) and QoL. The relationship between depression and diabetes mellitus was examined, given the significant difference in the prevalence of diabetes mellitus between AF and hypertensive patients. Presence of diabetes mellitus did not predict baseline depression in either group (odds ratio [OR], 0.28; 95% confidence interval [CI], ⫺ 4.26 to 4.83 [p ⫽ 0.902]; and OR, ⫺ 0.87; 95% CI, ⫺ 3.43 to 1.68 [p ⫽ 0.498] for AF and hypertensive patients, respectively) or for both groups combined (OR, ⫺ 1.27; 95% CI, ⫺ 3.36 to 0.82; p ⫽ 0.23). Higher levels of trait anxiety (mean score, p ⬍ 0.02) and percentage of scores ⱖ 40 on STAI, p ⫽ 0.03) were observed in the AF patients (Table 2). Depression and trait anxiety were highly comorbid conditions in AF patients, with 71% of patients reporting BDI scores ⱖ 10 also exhibiting high levels of anxiety. Female AF patients had a poorer QoL than male AF patients (t[99] ⫽ 3.43; p ⬍ 0.05), but no significant differences in any of the psychological parameters were observed between recurrent and permanent AF patients (data not shown). AF and hypertensive patients displaying significant symptoms of depression were demographically and clinically comparable to nondepressed patients
Table 2—Baseline and 6-Month Psychological Characteristics of the AF and Hypertensive Patients* Variables Baseline psychological characteristics Median (IQR) BDI score BDI score ⱖ 10 Mean (SD) trait anxiety score Trait anxiety score ⱖ 40 Mean (SD) QoL score 6-month psychological characteristics Median (IQR) BDI score BDI score ⱖ 10 Mean (SD) trait anxiety score Trait anxiety score ⱖ 40 Mean (SD) QoL score
AF Patients
Hypertensive Patients
p Value
7.0 (3–13)
6.0 (2–10)
0.07
38 (37.6) 37.4 (12.6)
29 (30.0) 33.3 (11.4)
0.32 0.02
38 (37.6)
21 (21.6)
0.03
20.4 (5.7)
19.9 (5.3)
0.57
7.0 (3–13)
6.0 (2–11)
0.61
28 (36.8) 36.9 (12.5)
20 (29.4) 32.6 (11.2)
0.48 0.03
24 (33.3)
17 (25.4)
0.25
20.3 (6.2)
20.5 (6.2)
0.95
*Data are presented as No. (%) unless otherwise indicated. IQR ⫽ interquartile range. CHEST / 132 / 4 / OCTOBER, 2007
1261
(p ⫽ not significant; data not shown). AF and hypertensive patients displaying elevated symptoms of trait anxiety were found to be younger (p ⫽ 0.02 and p ⬍ 0.01, respectively), with anxious hypertensive patients also more likely to be female (p ⬍ 0.01, data not shown). Predictors of QoL at Baseline Bivariate analysis indicated a strong association between depression and baseline QoL ( ⫽ 0.58, t ⫽ 9.78, p ⬍ 0.001); the higher the depression score the poorer the QoL. This association survived in the model also entering patient type and the interaction between depression and interaction score ( ⫽ 0.49), as well as adjustment for demographic factors ( ⫽ 0.50). Neither patient type ( ⫽ 0.03) nor the interaction ( ⫽ 0.10) were associated with QoL. Of the demographic variables, men and those in paid employment report better QoL than women ( ⫽ ⫺ 0.23) and those who were unemployed ( ⫽ 0.17). Trait anxiety was similarly associated with QoL ( ⫽ 0.56, t ⫽ 9.28, p ⬍ 0.001), and although the association was attenuated, it survived entry for patient type and the interaction term ( ⫽ 0.42) and demographic status ( ⫽ 0.46). Again, there were no differences between patient group in QoL ( ⫽ 0.04), nor was there an interaction between trait anxiety and patient type ( ⫽ 0.15). In a final one-step model in which depression and trait anxiety were entered, both emerged as significant predictors of contemporary QoL ( ⫽ 0.38, t ⫽ 3.99, p ⬍ 0.001; and  ⫽ 0.26, t ⫽ 2.77, p ⫽ 0.006, respectively). Persistence of Depression and Anxiety at 6 Months, and Stability of QoL The BDI and STAI were returned and completed at 6 months by 76 AF patients (75.2%) and 72 AF patients (71.3%), respectively. At 6 months, depression was present in 28 patients (36.8%) [Table 2], with 52.6% of patients reporting significant depression at baseline continuing to report elevated depressive symptoms at follow-up. Elevated levels of trait anxiety at follow-up were present in 24 patients (33.3%). Of the AF patients reporting high levels of anxiety at baseline, 52.6% continued to report elevated trait anxiety at 6 months. QoL at 6 months was strongly correlated with QoL at baseline (r ⫽ 0.67, p ⬍ 0.001), with the same association holding true for depression (r ⫽ 0.79, p ⬍ 0.001) and trait anxiety (r ⫽ 0.81, p ⬍ 0.001). QoL, depression, and anxiety did not change significantly between baseline and follow-up. The significant baseline difference in mean trait anxiety score 1262
between AF and hypertensive persisted at 6 months, although the difference in the dichotomized variable was no longer significant (p ⫽ 0.25). Predictors of QoL at 6 Months Outcomes were similar to those found at baseline. Depression at follow-up was strongly associated with contemporary QoL ( ⫽ 0.59, t ⫽ 8.52, p ⬍ 0.001). This association was not attenuated in a model adjusting for patient type and its interaction with depression ( ⫽ 0.62), and in a model additionally adjusting for demographic status ( ⫽ 0.65). Again, neither patient type ( ⫽ 0.04) nor the interaction ( ⫽ ⫺ 0.04) were associated with QoL. Of the demographic variables examined, gender ( ⫽ ⫺ 0.13) and employment status ( ⫽ 0.20) were related to QoL. Trait anxiety at 6 months was similarly related to 6-month QoL ( ⫽ 48, t ⫽ 6.34, p ⬍ 0.001). Again, the regression coefficients remained high in subsequent models adjusting for patient status and demographics ( ⫽ 0.39 in both cases). In contrast to the baseline findings, only depression emerged as a predictor of QoL in the model that entered both depression ( ⫽ 0.60, t ⫽ 4.94, p ⬍ 0.001) and trait anxiety ( ⫽ 0.00). Predictors of the Change in QoL Over 6-Month Follow-up Change in QoL between baseline and 6 months was significantly related to change in depression score ( ⫽ 0.37, t ⫽ 4.68, p ⬍ 0.001); improvements in depression were associated with enhanced QoL. This association was not influenced by patient type ( ⫽ 0.36) nor by baseline demographic characteristics ( ⫽ 0.40). Neither patient type nor demographic status was related to the direction and extent of change in QoL. The same was true for change in trait anxiety ( ⫽ 0.35, t ⫽ 4.22, p ⬍ 0.001); in this case, the association was abolished in models that included patient type and demographic status.
Discussion Depression and anxiety have been found to be strong predictor of QoL in myocardial infarction patients.17 Although QoL has been extensively studied in patients with AF,8 depression and anxiety have yet to receive attention. There has been no previous examination of the association of the affective status of AF patients and their QoL. Rates of depression and trait anxiety observed in our AF cohort were comparable to that seen in patients following myocardial infarction,13,18 –22 imOriginal Research
plying similar levels of psychological morbidity characterize these two patient groups. In the present study, hypertensive patients presented with similar levels of depression to patients with AF, although they exhibited lower levels of trait anxiety. In contrast to AF patients, the prevalence of depression in hypertensive patients has received some attention.23–25 Rabkin et al24 demonstrated that major depression was three times more common among hypertensive patients even when known cardiac risk factors were taken into account. Mild depressive symptoms are present in 20% of hypertensive patients.25 Significant symptoms of depression and trait anxiety persisted over the first 6 months of follow-up (Table 2), with approximately one half of all patients reporting elevated symptoms of depression and anxiety at baseline, also reporting symptoms similarly increased at 6 months. In addition, the average depression and trait anxiety scores showed no attenuation from baseline to follow-up. Such observations suggest that AF patients experience significant psychological morbidity, which is protracted and not limited to a single point of observation. The current study reveals no significant baseline differences in QoL between patients with AF and hypertensive patients in sinus rhythm. We are only aware of one previous study25 that examined QoL in AF patients and a hypertensive control group, which reported a lower QoL on six of the eight scales of the Short Form-36 (SF-36) in AF patients compared to hypertensive control subjects. It is plausible that this apparent discrepancy in results arises from the exaggerated baseline scores in the previous study,26 due to the highly symptomatic patients undergoing an invasive procedure (pulmonary vein isolation). Differences in QoL between patients with recurrent AF and permanent AF have not been subject to investigation. The current study, although not intentionally powered to detect differences between the subgroups of AF patients, demonstrated that patients with paroxysmal AF showed comparable levels of QoL to patients with permanent AF. Only one of the five observational studies4 –7,27 previously examining QoL did so exclusively in a cohort of permanent AF patients, and found that men with permanent AF displayed similar levels of QoL (SF-36) to an aged-matched control group in sinus rhythm.5 The authors argued that having a predictable clinical course of rate control and anticoagulation, in addition to relatively well-controlled heart rates, might explain the equivocal findings. However, one study28 demonstrated that there was no significant association between achieved heart rate either at rest or during exercise and QoL. Conversely, the significant impairment in QoL reported in paroxysmal AF www.chestjournal.org
patients4 – 6,27 may be a consequence of rapid heart rates that are potentially more symptom producing than permanent (rate-controlled) AF, or individuals perceiving their illness to be more intrusive on their everyday lives.29 Female AF patients had a poorer QoL than men. Previous research on gender differences in QoL in AF patients demonstrated that female patients display poorer physical component summary scores on the SF-36 than their male counterparts.27 Although this result was confounded by the fact that female patients were significantly older than male patients, such discrepancies may not fully account for gender differences in QoL.27 Indeed, the greater impact of AF on QoL in women may be attributed to their heightened sensitivity to the disease and its associated symptoms, gender differences in the perception of illness, or a lower threshold for reporting illness burden.27 Increased sensitivity to symptoms is unlikely to be a major factor driving these differences because the impairment of QoL experienced by women persisted even after controlling for somatization scores.27 Depression and anxiety were strongly associated with QoL in our cross-sectional analysis. However, our QoL measure did not include the affective item in the Dartmouth COOP chart system. Further, the associations emerged at both baseline and follow-up, were not influenced by patient type, and survived adjustment for a number of sociodemographic variables. Whereas at baseline, depression and anxiety appeared to be independently related to QoL, at 6 months only depression emerged from analysis entering both independent variables. This resonates with the results from a study13,17 of QoL in myocardial infarction patients, in which QoL is more consistently predicted by depression than by anxiety. The intimacy of the association between depression and QoL is best demonstrated in the present study by the change score analysis. Changes in depression over 6 months were significantly associated with changes in QoL, even taking into account possible confounding variables; this was not the case for anxiety. This holds strong implications for the psychological management of AF patients. Our results suggest that depression should be the focus because reductions in depression should translate into improvements in QoL. Increasingly, QoL is regarded as an important clinical outcome in the management of chronic conditions, and the present study suggests a strategy for improving QoL in AF patients may be needed. In conclusion, approximately one third of AF patients have elevated levels of depression and anxiety, which persist at 6 months. Symptoms of depresCHEST / 132 / 4 / OCTOBER, 2007
1263
sion were the strongest independent predictor of future QoL in these patients. References 1 Bilal IM, Taneja AK, Lip GY, et al. Recent developments in atrial fibrillation. BMJ 2005; 330:238 –243 2 Dewilde S, Carey IM, Emmas C, et al. Trends in the prevalence of diagnosed atrial fibrillation, its treatment with anticoagulation, and predictors of such treatment in UK primary care. Heart 2006; 92:1064 –1070 3 Ruigomez A, Johansson S, Wallander MA, et al. Incidence of chronic atrial fibrillation in general practice and its treatment patterns. J Clin Epidemiol 2002; 55:358 –363 4 Dorian P, Jung W, Newman D, et al. The impairment of health-related quality of life in patients with intermittent atrial fibrillation: implications for assessment of investigational therapy. J Am Coll Cardiol 2000; 36:1303–1309 5 Howes CJ, Reid MC, Brandt C, et al. Exercise tolerance and quality of life in elderly patients with chronic atrial fibrillation. J Cardiovasc Pharmacol Ther 2001; 6:23–29 6 van den Berg MP, Hassink RJ, Tuinenburg AE, et al. Quality of life in patients with paroxysmal atrial fibrillation and its predictors: importance of the autonomic system. Eur Heart J 2001; 22:247–253 7 Kang Y, Bahler R. Health related quality of life in patients newly diagnosed with atrial fibrillation. Eur J Cardiovasc Nurs 2003; 3:71–76 8 Thrall G, Lane D, Carroll D, et al. Quality of life in patients with atrial fibrillation: a systematic review. Am J Med 2006; 119:448.e1– e19 9 Townsend P. Deprivation. J Soc Policy 1987; 16:125–146 10 Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561–571 11 Spielberger CD, Gorsuch RL, Lushene R, et al. State-Trait Anxiety Inventory for adults. Palo Alto, CA: Consulting Psychologists Press, 1983 12 Nelson E, Wasson J, Kirk J, et al. Assessment of function in routine clinical practice: description of the COOP chart method and preliminary findings. J Chron Dis 1987; 40:55S– 63S 13 Lane D, Carroll D, Ring C, et al. Mortality and quality of life 12 months after myocardial infarction: effects of depression and anxiety. Psychosom Med 2001; 63:221–230 14 Frasure-Smith N, Lesperance F, Gravel G, et al. Depression and health-care costs during the first year following myocardial infarction. J Psychosom Res 2000; 48:471– 478
1264
15 Frasure-Smith N, Lesperance F, Talajic M. Depression and 18-month prognosis after myocardial infarction. Circulation 1995; 91:999 –1005 16 Akin LS, West SG. Multiple regression: testing and interpreting interactions. London, UK: Sage, 1991 17 Lane D, Carroll D, Ring C, et al. Effects of depression and anxiety on mortality and quality-of-life 4 months after myocardial infarction. J Psychosom Res 2000; 49:229 –238 18 Frasure-Smith N, Lesperance F, Talajic M. The impact of negative emotions on prognosis following myocardial infarction: is it more than depression? Health Psychol 1995; 14:388 –398 19 Frasure-Smith N, Lesperance F, Juneau M, et al. Gender, depression, and one-year prognosis after myocardial infarction. Psychosom Med 1999; 61:26 –37 20 Lane D, Carroll D, Ring C, et al. In-hospital symptoms of depression do not predict mortality 3 years after myocardial infarction. Int J Epidemiol 2002; 31:1179 –1182 21 Ahern DK, Gorkin L, Anderson JL, et al. Biobehavioral variables and mortality or cardiac arrest in the Cardiac Arrhythmia Pilot Study (CAPS). Am J Cardiol 1990; 66: 59 – 62 22 Denollet J, Brutsaert DL. Personality, disease severity, and the risk of long-term cardiac events in patients with a decreased ejection fraction after myocardial infarction. Circulation 1998; 97:167–173 23 Scalco AZ, Scalco MZ, Azul JB, et al. Hypertension and depression. Clinics 2005; 60:241–250 24 Rabkin JG, Charles E, Kass F. Hypertension and DSM-III depression in psychiatric outpatients. Am J Psychiatry 1983; 140:1072–1074 25 Bonnet F, Irving K, Terra JL, et al. Depressive symptoms are associated with unhealthy lifestyles in hypertensive patients with the metabolic syndrome. J Hypertens 2005; 23:611– 617 26 Nilsson B, Chen X, Svendsen JH. Effects of pulmonary vein isolation on quality of life in patients with paroxysmal atrial fibrillation. Heart Drug 2003; 3:173–179 27 Paquette M, Roy D, Talajic M, et al. Role of gender and personality on quality-of-life impairment in intermittent atrial fibrillation. Am J Cardiol 2000; 86:764 –768 28 Cooper HA, Bloomfield DA, Bush DE, et al. Relationship between achieved heart rate and outcomes in patients with atrial fibrillation (AFFIRM Study). Am J Cardiol 2004; 93:1247–1253 29 Devins GM. Illness intrusiveness and the psychosocial impact of lifestyle disruptions in chronic life-threatening disease. Adv Ren Replace Ther 1994; 1:251–263
Original Research