Abstract
Congestive heart failure (CHF) is the most important cause of death in patients with atrial fibrillation (AF). We aimed to study the association between cardiovascular drugs in AF patients and incident CHF. The study population included all adults (n = 120 756) aged ≥45 years diagnosed with AF in Sweden diagnosed for the period 1998‐2006. Outcome was incident congestive heart failure (follow‐up 2007‐2015) in AF patients. Associations between treatment with cardiovascular pharmacotherapies and CHF were evaluated using Cox regression to estimate hazard ratios (HRs) with 95% CIs, after adjustment for age, sociodemographic variables, and comorbidities. During a mean 5.3 years (SD 3.0) of follow‐up, there were 28 257 (23.4%) incident cases of CHF. Treatment with beta‐1‐selective and non‐selective beta‐blockers and statins was associated with lower risks of incident CHF in men, HR, (95% CI); 0.90, (0.87‐0.94); 0.90, (0.84‐0.97), and 0.94, (0.90‐0.99), respectively. Only beta‐1‐selective beta‐blockers were protective in women 0.94 (0.91‐0.98). Treatment with loop diuretics, potassium‐saving agents, ACE inhibitors, and angiotensin receptor blockers was associated with a higher risk of CHF. For men, treatment with heart‐active calcium channel blockers also led to a higher risk of CHF. In conclusion, we found that beta‐blockers, in particular, but also statins were associated with lower risk of incident CHF in patients with AF.
Keywords: atrial fibrillation, congestive heart failure, drug treatment, gender
1. INTRODUCTION
Atrial fibrillation (AF) is the most common arrhythmia in the world, with an estimated prevalent number of individuals with AF of 46 million. 1 The prevalence is estimated to be around 2% among adults aged 20 years and above in Europe, 2 with the highest prevalence in Sweden of almost 3%. 3 , 4
Even if ischemic stroke is the most well‐known disease associated with AF, 5 , 6 congestive heart failure (CHF) is also of great importance and is three times more common in AF patients compared with non‐AF patients. 7 CHF and AF are interrelated diseases 8 , 9 with one that could precede the other. Both stroke 10 and CHF 7 are significant contributors to mortality in patients with AF.
As a registered diagnosis of hypertension in some earlier studies has been shown previously to be associated with a lower risk of CHF in AF patients, 11 studies on antihypertensive drugs in relation to incident CHF in AF patients are of interest. However, most studies analyzing the association between different antihypertensive drugs and incident CHF are performed among patients with hypertension and showed an effect via diuretic. 12 In a study, more specifically on patients with AF thiazides, vessel‐active calcium channel blockers (CCBs), and non‐selective beta‐blockers were associated with a lower risk of incident CHF, 13 but this study was performed on AF patients in a primary care setting. Thus, it is of interest to study the effect of antihypertensive drugs in a larger, national cohort of AF patients.
The aim was to study cardiovascular pharmacotherapies and the associated risk of incident CHF in men and women with AF. We also aimed to explore this in different age‐groups, and in men and women with and without prevalent coronary heart disease (CHD).
2. METHODS
2.1. Design and study population
This study used individual‐level patient data from the Swedish National Patient Register (NPR) with nationwide coverage of hospitalizations since 1987, including diagnoses from hospital ambulatory care but not from primary care. Personal identification numbers were replaced with new unique serial numbers to ensure anonymity. The files were linked to the Total Population Register and the Swedish Cause of Death Register, which contains individual‐level data on age, sex, education, and hospital admissions for all residents registered in Sweden.
The study included patients that had been diagnosed with AF as identified by the presence of the ICD‐10 code in any position (10th version of the WHO's International Classification of Diseases) for atrial fibrillation (I48) in the Swedish NPR, with a total of 167 166 patients identified during 1998 through 2006, aged 45 years or above and alive on January 1, 2007. Patients with hospital stays for CHF between January 1, 1998, and December 31, 2006, were excluded (n = 43 479), as well as patients with one dispensed prescription between July 1, 2005, until December 31, 2006 (n = 2931). Included in the study were the following: 120 756 individuals (67 656 men and 53 100 women) with AF, among whom 28 257 (15 093 men and 13 164 women) were hospitalized for CHF during follow‐up, aged 45 years or above on January 1, 2007.
2.2. Outcome variable, pharmacotherapies, sociodemographic variables, and comorbidities
Time to first CHF episode during the study period (from January 1, 2007, until end of follow‐up, December 31, 2015) was defined as having a diagnosis of CHF ICD‐10 code (I50 or I110) in the Swedish NPR.
Data on dispensed prescriptions of drugs, classified according to the Anatomic Therapeutic Chemical (ATC) Classification, were collected from the Swedish Prescribed Drug Register (National Board of Health and Welfare) from July 1, 2005 (ie, start of the register), until December 31, 2006, with dispensed prescriptions on at least two occasions. 14 , 15 We reckoned that 18 months would be sufficient to assess whether a person was prescribed a certain drug or not. Substances were recorded, meaning that medical drugs could be single‐substance drugs, or combined medical drugs with the substances recorded separately, for example, a RAS‐blocking agent combined with a thiazide, which would thus be recorded into two separate groups. Digitalis agents (C01AA) were recorded. Diuretic drugs (C03) were recorded as loop diuretics (C03C), thiazides or related agents, also registered when in combination with other drugs (C03A, C03B, C03E, C09B, and C09DA), potassium‐saving agents (C03DA), beta‐1‐selective agents (C07AB and C07FB), non‐selective agents beta‐blockers (C07AA and C07AG), heart‐active agents CCBs (C08DA), vessel‐active CCBs (C08CA and C08DB), ACE inhibitors (C09AA and C09BA), angiotensin receptor blocking agents (ARB) (C09CA, C09DA, and C09DB), and statins (C10AA).
Demographic and socioeconomic variables: Sex: Men and women. Individuals were divided into the following age‐groups: 45‐64, 65‐84, and ≥85 years. Individuals younger than 45 years were excluded as they rarely develop AF and CHF Educational level was categorized as ≤9 years (partial or complete compulsory schooling), 10‐12 years (partial or complete secondary schooling), and >12 years (college and/or university studies). Marital status was classified as married, unmarried, divorced, or widowed. Information on marital status was missing for 14 subjects. The neighborhood socioeconomic status (SES) areas were categorized into three groups according to the neighborhood index: more than one standard deviation (SD) below the mean (high SES or low deprivation level), more than one SD above the mean (low SES or high deprivation level), and within one SD of the mean (middle SES or deprivation level).
We identified the following cardiovascular comorbidities among individuals in the study population: hypertension (I10‐15), CHD (I20‐25), valvular heart diseases (I05‐08 and I34‐38), cardiomyopathy (I42), and cerebrovascular diseases (CVDs; I60‐69). In addition, we also included obesity (E65‐E68), diabetes mellitus (E10‐14), COPD (J40‐J47), obstructive sleep apnea syndrome (G47), depression (F32‐F34, F38‐F39), and anxiety disorders (F40‐41). All comorbidities were assessed prior to Jan 2007 at the same time point as the index AF event.
2.3. Statistical analyses
Cox regression with hazard ratios (HRs) and 95% confidence interval (95% CI) using time to first diagnosis of CHF as the outcome was used. The proportional hazards assumptions were evaluated by log‐log curves. The following were studied in the analyses: loop diuretics, thiazides, aldosterone antagonists, beta‐1‐selective beta‐blockers, unselective beta‐blocker, heart‐active CCBs, vessel‐active CCBs, ACE inhibitors, and angiotensin receptor blockers (ARB). In the full model, we adjusted for age, region of Sweden, immigrant status, socioeconomic factors, comorbidities, and all drug classes. Furthermore, we also studied the most used specific drugs of the following groups: beta‐1‐selective beta‐blockers, unselective beta‐blockers, heart‐active CCBs, vessel‐active CCBs, ACE inhibitors, and angiotensin receptor blockers (ARB). We also stratified Cox regression into subgroups by age, subjects aged 45‐64, 65‐84 and ≥85 years of age, and into those with CHD vs those without CHD.
As a sensitivity analysis, we only included patients with a recorded AF diagnosis during 2005 and 2006, with a Cox regression analysis adjusted in the same way as the original analysis in men and women. Furthermore, we also performed another sensitivity analysis with incident CHF as outcome categorizing patients into those with heart disease (CHD, cardiomyopathy, or valvular heart disease) or without. We also analyzed the association between the dispensed drug groups and mortality in heart diseases (CHD, cardiomyopathy, or valvular heart disease) and performed sensitivity analyses for all cardiovascular diseases and for stroke. In further analyses, we analyzed the proportion of dispensed drugs over time, the percentage of multiple dispensed drugs in the study sample as well as in patients with incident CHF, and finally, we used a competing risk model as a sensitivity analysis.
A two‐sided P‐value of <.05 was considered statistically significant for variables in the Cox regression. All analyses were performed in SAS 9.4.
3. RESULTS
Table 1 shows the characteristics of the study population of 120 756 individuals with a mean age of 74.9 years (SD 11.0) (67 656 men and 53 100 women, with mean ages (± SD) 72.5 ± 10.9 for men, and 78.0 ± 10.3), 28 257 (23.4%) with a recorded episode of CHF (15 093 men, 22.3%, and 13 164 women, 24.8%). The mean follow‐up time was 5.3 years (SD 3.0), 5.5 years (SD 3.0) for men, and 5.0 years (SD 3.1) for women, and HRs were calculated based on 636 381 person‐years at risk; 373 039 among men and 263 342 among women. Table 1 also shows rates of dispensed pharmacological drugs in subjects with or without incident CHF diagnosis during follow‐up.
TABLE 1.
Data for patients aged 45 y and older with diagnosed atrial fibrillation (AF) (1998‐2006), and also for AF patients with newly diagnosed heart failure (2007‐2015) in Sweden
| Total population | Heart failure events | |||
|---|---|---|---|---|
| No. | % | No. | % | |
| Total population | 120 756 | 28 257 | 23 | |
| Gender | ||||
| Males | 67 656 | 56.0 | 15 093 | 53.4 |
| Females | 53 100 | 44.0 | 13 164 | 46.6 |
| Age (y) | ||||
| 45‐64 | 23 303 | 19.3 | 2570 | 9.1 |
| 65‐84 | 71 963 | 59.6 | 18 691 | 66.1 |
| 85+ | 25 490 | 21.1 | 6996 | 24.8 |
| Educational attainment | ||||
| ≤9 years | 56 761 | 47.0 | 14 657 | 51.9 |
| 10‐12 years | 28 646 | 23.7 | 6583 | 23.3 |
| >12 years | 35 349 | 29.3 | 7017 | 24.8 |
| Immigrant status | ||||
| Sweden | 108 839 | 90.1 | 25 448 | 90.1 |
| Others | 11 917 | 9.9 | 2809 | 9.9 |
| Marital status | ||||
| Married | 66 987 | 55.5 | 14 159 | 50.1 |
| Not married/Widowed/Divorced | 53 769 | 44.5 | 14 098 | 49.9 |
| Neighborhood deprivation | ||||
| Low | 14 459 | 12.0 | 3063 | 10.8 |
| Middle | 63 030 | 52.2 | 15 197 | 53.8 |
| High | 14 429 | 11.9 | 3780 | 13.4 |
| Unknown | 28 838 | 23.9 | 6217 | 22.0 |
| Comorbidities: | ||||
| Hypertension | 34 121 | 28.3 | 8968 | 31.7 |
| CHD | 27 954 | 23.1 | 8652 | 30.6 |
| Stroke | 19 871 | 16.5 | 4764 | 16.9 |
| Obesity | 964 | 0.8 | 292 | 1.0 |
| Diabetes | 11 598 | 9.6 | 3552 | 12.6 |
| COPD | 6614 | 5.5 | 2112 | 7.5 |
| Depression | 2771 | 2.3 | 616 | 2.2 |
| Anxiety | 1633 | 1.4 | 338 | 1.2 |
| Alcoholism | 2236 | 1.9 | 481 | 1.7 |
| Sleep apnea | 2527 | 2.1 | 631 | 2.2 |
| Cardiomyopathy | 1305 | 1.1 | 539 | 1.9 |
| Valvular heart disease | 7613 | 6.3 | 2712 | 9.6 |
| Cardiovascular medication | 106 537 | 88.2 | 26 603 | 94.1 |
| a. Loop diuretics (C03C) | 7524 | 6.2 | 2137 | 7.6 |
| b. Thiazides (C03A, C03B, C03E, C09B C09DA) | 6004 | 5.0 | 1455 | 5.1 |
| c. Potassium‐saving agents (C03D) | 1992 | 1.6 | 686 | 2.4 |
| d. Beta‐1‐selective beta‐blockers (C07AB and C07FB) and | 35 969 | 29.8 | 7798 | 27.6 |
| e. Non‐selective beta‐blockers (C07AA and C07AG), with | 7577 | 6.3 | 1536 | 5.4 |
| f. Heart‐active calcium channel blockers (C08DA) and | 1932 | 1.6 | 482 | 1.7 |
| g. Vessel‐active calcium channel blockers (C08CA and C08DB) | 10 444 | 8.6 | 2818 | 10.0 |
| h. ACE inhibitors (C09AA and C09BA) and | 15 377 | 12.7 | 4566 | 16.2 |
| i. Angiotensin receptor blocking (ARB) agents (C09CA, C09DA, and C09DB) | 7197 | 6.0 | 2039 | 7.2 |
| j. Statins (C10AA) | 15 004 | 12.4 | 3714 | 13.1 |
Abbreviations: CHD, Coronary heart disease; COPD, chronic obstructive pulmonary disease.
Cox regression models are shown in Table 2 for men and women, respectively, with incident CHF as outcome, also divided into subgroups of age. HRs were fairly consistent over time. In men, as well as in the age‐group 65‐84 years, non‐selective beta‐blockers and statins were associated with overall lower risks of CHF. In women, non‐selective beta‐blockers were not associated with a lower risk and statins were only associated with a lower risk of CHF in the age‐group 65‐84 years. For both men and women, treatment with beta‐1‐selective beta‐blockers was associated with a lower risk of CHF in all age‐groups for men, but only in the age‐groups 65‐84 years among women. For both men and women, treatment with loop diuretics, potassium‐saving agents, ACE inhibitors, and ARBs was associated with higher risk of CHF, in all age‐groups, except for ARBs among men 85 years of age and above, and for ARBs among women 45‐64 years. In men, non‐selective beta‐blockers and statins were associated with lower risks of CHF, for non‐selective beta‐blockers also in all age‐groups, and for statins in the age‐group 65‐84 years old. In women, treatment with heart‐active calcium receptor blockers was associated with a higher risk of CHF, also found in the age‐groups 85 years of age and older. The sensitivity analysis of patients (n = 73 451) diagnosed during the years 2005‐2006 showed similar results (Table S1). Among men, statins were not associated with lower risk (HR 0.97, 95% CI 0.92‐1.03) and, among women, thiazides were associated with a lower risk (HR 0.89, 95% CI 0.811‐0.98). Finally, results for heart‐active CCBs among women were attenuated and no longer significant (HR 1.14, 95% CI 0.97‐1.33).
TABLE 2.
The association between dispensed drugs and the risk of heart failure in men and women of different age‐groups with atrial fibrillation a
| All ages | 45‐64 y | 65‐84 y | 85+ y | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | HR | 95% CI | HR | 95% CI | |||||
| Men | ||||||||||||
| All medication | 1.59 | 1.49 | 1.69 | 1.66 | 1.43 | 1.93 | 1.48 | 1.37 | 1.61 | 1.76 | 1.53 | 2.02 |
| a. Loop diuretics | 1.39 | 1.30 | 1.49 | 2.43 | 1.81 | 3.26 | 1.51 | 1.37 | 1.65 | 1.28 | 1.15 | 1.44 |
| b. Thiazides | 1.03 | 0.95 | 1.11 | 1.15 | 0.91 | 1.45 | 1.01 | 0.92 | 1.11 | 1.00 | 0.85 | 1.19 |
| c. Potassium‐saving agents | 1.53 | 1.35 | 1.74 | 1.97 | 1.24 | 3.14 | 1.53 | 1.29 | 1.81 | 1.50 | 1.20 | 1.88 |
| d. Beta‐1‐selective beta‐blockers | 0.90 | 0.87 | 0.94 | 0.96 | 0.86 | 1.06 | 0.91 | 0.87 | 0.95 | 0.86 | 0.79 | 0.94 |
| e. Non‐selective beta‐blockers | 0.90 | 0.84 | 0.97 | 0.86 | 0.71 | 1.05 | 0.91 | 0.83 | 0.99 | 0.90 | 0.76 | 1.08 |
| f. Heart‐active calcium channel blockers | 1.07 | 0.94 | 1.23 | 1.09 | 0.72 | 1.64 | 1.02 | 0.86 | 1.20 | 1.25 | 0.94 | 1.65 |
| g. Vessel‐active calcium channel blockers | 1.09 | 1.03 | 1.15 | 1.08 | 0.91 | 1.28 | 1.08 | 1.01 | 1.15 | 1.08 | 0.96 | 1.22 |
| h. ACE inhibitors | 1.24 | 1.19 | 1.29 | 1.42 | 1.26 | 1.60 | 1.19 | 1.13 | 1.25 | 1.27 | 1.15 | 1.40 |
| i. ARB | 1.17 | 1.10 | 1.24 | 1.26 | 1.05 | 1.50 | 1.16 | 1.08 | 1.25 | 1.14 | 0.97 | 1.34 |
| j. Statins | 0.94 | 0.90 | 0.99 | 0.90 | 0.79 | 1.03 | 0.93 | 0.88 | 0.98 | 0.97 | 0.84 | 1.11 |
| Women | ||||||||||||
| All medication | 2.03 | 1.85 | 2.22 | 2.37 | 1.76 | 3.20 | 1.73 | 1.54 | 1.94 | 2.25 | 1.92 | 2.64 |
| a. Loop diuretics | 1.21 | 1.14 | 1.28 | 1.98 | 1.43 | 2.74 | 1.43 | 1.31 | 1.56 | 1.10 | 1.01 | 1.20 |
| b. Thiazides | 0.94 | 0.87 | 1.01 | 0.74 | 0.48 | 1.13 | 0.98 | 0.89 | 1.07 | 0.88 | 0.77 | 1.00 |
| c. Potassium‐saving agents | 1.31 | 1.19 | 1.44 | 2.42 | 1.46 | 4.00 | 1.42 | 1.25 | 1.61 | 1.17 | 1.01 | 1.35 |
| d. Beta‐1‐selective beta‐blockers | 0.94 | 0.91 | 0.98 | 0.98 | 0.84 | 1.15 | 0.92 | 0.88 | 0.97 | 0.96 | 0.90 | 1.03 |
| e. Non‐selective beta‐blockers | 0.96 | 0.89 | 1.03 | 0.91 | 0.68 | 1.22 | 0.94 | 0.86 | 1.03 | 0.98 | 0.85 | 1.13 |
| f. Heart‐active calcium channel blockers | 1.15 | 1.02 | 1.30 | 0.89 | 0.52 | 1.52 | 1.12 | 0.95 | 1.31 | 1.23 | 1.00 | 1.51 |
| g. Vessel‐active calcium channel blockers | 1.02 | 0.96 | 1.08 | 0.98 | 0.72 | 1.33 | 1.00 | 0.93 | 1.08 | 1.05 | 0.95 | 1.17 |
| h. ACE inhibitors | 1.24 | 1.18 | 1.30 | 1.53 | 1.22 | 1.91 | 1.19 | 1.12 | 1.27 | 1.27 | 1.16 | 1.39 |
| i. ARB | 1.15 | 1.07 | 1.23 | 1.06 | 0.78 | 1.42 | 1.11 | 1.02 | 1.21 | 1.22 | 1.08 | 1.37 |
| j. Statins | 0.97 | 0.91 | 1.02 | 1.06 | 0.85 | 1.33 | 0.93 | 0.87 | 0.99 | 0.95 | 0.84 | 1.08 |
In the full model, we adjusted for age, region of Sweden, immigrant status, socioeconomic factors, comorbidities, and all drug classes.
Heart diseases include: CHD, cardiomyopathy, and valvular heart diseases (including rheumatic and non‐rheumatic valvular heart diseases).
Abbreviations: ARBs, angiotensin receptor blocking agents; CI, confidence intervals; HR, hazards ratios.
Full adjusted model.
Bold values are statistically significant.
Analyses of specific drugs (Table 3) showed significantly lower risk for beta‐1‐selective beta‐blockers, that is, for metoprolol among men and women and, for non‐selective beta‐blockers, for sotalol among men and pindolol among women. Significantly higher risks were found for heart‐active CCBs, that is, verapamil among women, for vessel‐active CCBs, that is, amlodipine among men, for ACE inhibitors, that is, enalapril and ramipril for both men and women, for all others drugs for men, for ARBs, that is, losartan among men and women, and for candesartan in women.
TABLE 3.
The association between specific dispensed drugs and the risk of heart failure in patients with atrial fibrillation a
| (No. With/without heart failure) | Men | Women | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | |||
| Beta‐1‐selective beta‐blockers | ||||||
| Metoprolol (3473/12 942) | 0.86 | 0.82 | 0.91 | 0.93 | 0.89 | 0.98 |
| Atenolol (2698/9353) | 0.96 | 0.91 | 1.02 | 0.97 | 0.92 | 1.02 |
| Others (1541/5603) | 0.99 | 0.92 | 1.06 | 0.97 | 0.90 | 1.04 |
| Non‐selective beta‐blockers | ||||||
| Sotalol (1193/4884) | 0.84 | 0.78 | 0.92 | 0.96 | 0.88 | 1.04 |
| Pindolol (203/913) | 0.94 | 0.76 | 1.16 | 0.81 | 0.67 | 0.97 |
| Heart‐active calcium receptor blockers | ||||||
| Verapamil (482/1450) | 1.07 | 0.94 | 1.23 | 1.15 | 1.02 | 1.30 |
| Diltiazem (105/268) | 1.12 | 0.85 | 1.48 | 1.22 | 0.94 | 1.60 |
| Vessel‐active calcium receptor blockers | ||||||
| Amlodipine (1172/2949) | 1.12 | 1.03 | 1.21 | 1.09 | 0.99 | 1.19 |
| Felodipine (1411/4037) | 1.04 | 0.96 | 1.11 | 0.97 | 0.89 | 1.05 |
| Others (130/372) | 1.18 | 0.93 | 1.49 | 0.95 | 0.74 | 1.23 |
| ACE inhibitors | ||||||
| Enalapril (2922/7362) | 1.19 | 1.13 | 1.25 | 1.19 | 1.12 | 1.26 |
| Ramipril (1273/2525) | 1.27 | 1.18 | 1.37 | 1.34 | 1.22 | 1.47 |
| Others (302/761) | 1.18 | 1.02 | 1.36 | 1.10 | 0.91 | 1.32 |
| ARBs | ||||||
| Losartan (833/1799) | 1.28 | 1.16 | 1.40 | 1.23 | 1.11 | 1.36 |
| Candesartan (649/1764) | 1.10 | 0.98 | 1.23 | 1.12 | 1.01 | 1.25 |
| Others (557/1595) | 1.08 | 0.97 | 1.21 | 1.04 | 0.91 | 1.18 |
Abbreviations: ARBs, angiotensin receptor blocking agents; HR, hazards ratios; CI, confidence intervals.
Full adjusted model.
Bold values are statistically significant.
The association between cardiovascular pharmacotherapies and risk of CHF stratified for the presence or absence of CHD are shown in Table 4 for men and women, respectively. Lower risks of CHF were found among men, with or without CHD, treated with beta‐1‐selective beta‐blockers while among women this association was found only for those without CHD. A lower CHF risk was also found for men treated with non‐selective beta‐blockers. Treatment with thiazides was associated with a lower risk of CHF among women without CHD. Men with CHD, who were treated with statins, also had a lower risk of heart failure during the follow‐up. Higher risks for CHF were found for men and women, with and without CHD, treated with loop diuretics, ACE inhibitors or ARBs and for men treated with potassium‐saving agents. Higher risk for CHF was found among men with CHD treated with vessel‐active CCBs and among women without CHD treated with potassium‐saving agents or heart‐active CCBs.
TABLE 4.
The association between dispensed drugs and the risk of heart failure in patients with atrial fibrillation in patients with and without coronary heart disease (CHD) a
| Men | Women | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| With CHD | Without CHD | With CHD | Without CHD | |||||||||
| HR | 95% CI | HR | 95% CI | HR | 95% CI | HR | 95% CI | |||||
| All medication | 1.53 | 1.27 | 1.84 | 1.58 | 1.48 | 1.69 | 2.93 | 2.09 | 4.12 | 1.95 | 1.77 | 2.14 |
| a. Loop diuretics | 1.23 | 1.07 | 1.41 | 1.45 | 1.34 | 1.57 | 1.25 | 1.11 | 1.41 | 1.19 | 1.11 | 1.28 |
| b. Thiazides | 0.92 | 0.78 | 1.08 | 1.06 | 0.97 | 1.15 | 1.03 | 0.88 | 1.22 | 0.91 | 0.84 | 0.99 |
| c. Potassium‐saving agents | 1.73 | 1.38 | 2.17 | 1.44 | 1.23 | 1.68 | 1.07 | 0.87 | 1.32 | 1.38 | 1.24 | 1.53 |
| d. Beta‐1‐selective beta‐blockers | 0.90 | 0.84 | 0.96 | 0.92 | 0.88 | 0.96 | 0.93 | 0.87 | 1.01 | 0.95 | 0.91 | 0.99 |
| e. Non‐selective beta‐blockers | 0.97 | 0.83 | 1.13 | 0.89 | 0.82 | 0.97 | 0.90 | 0.76 | 1.07 | 0.97 | 0.90 | 1.06 |
| f. Heart‐active calcium channel blockers | 0.98 | 0.70 | 1.36 | 1.10 | 0.95 | 1.27 | 1.14 | 0.84 | 1.53 | 1.14 | 1.00 | 1.31 |
| g. Vessel‐active calcium channel blockers | 1.12 | 1.03 | 1.22 | 1.06 | 0.99 | 1.14 | 1.02 | 0.91 | 1.13 | 1.02 | 0.95 | 1.10 |
| h. ACE inhibitors | 1.15 | 1.07 | 1.23 | 1.28 | 1.22 | 1.35 | 1.13 | 1.03 | 1.24 | 1.28 | 1.21 | 1.36 |
| i. ARBs | 1.14 | 1.02 | 1.27 | 1.18 | 1.09 | 1.27 | 1.14 | 1.00 | 1.29 | 1.16 | 1.07 | 1.25 |
| j. Statins | 0.91 | 0.85 | 0.97 | 0.96 | 0.90 | 1.03 | 0.95 | 0.88 | 1.04 | 0.96 | 0.89 | 1.04 |
In the full model, we adjusted for age, region of Sweden, immigrant status, socioeconomic factors, comorbidities, and all drug classes.
Abbreviations: ARBs, angiotensin receptor blocking agents; HR, hazards ratios, CI, confidence intervals.
Full adjusted model.
Bold values are statistically significant.
The analysis of patients with AF with or without heart diseases (CHD, cardiomyopathy, or valvular heart disease) showed similar results (Table S2).
Analyses of mortality in heart diseases are shown in Table 5. There was an increased mortality rate found for treatment with loop diuretics or ACE inhibitors for both men and women, and for men also for potassium‐saving agents and heart‐active CCBs. A decreased mortality rate was found for treatment with beta‐1‐selective beta‐blockers, non‐selective beta‐blockers, and statins. As sensitivity analyses, mortality in all cardiovascular diseases and in stroke was specifically studied (Tables S3 and S4). Similar results for cardiovascular mortality were found as for mortality in heart diseases, but the mortality for heart‐active CCBs was not significant (HR 1.12, 95% CI 0.99‐1.27). An increased stroke mortality was found for treatment with loop diuretics among men, treatment with vessel‐active CCBs among men and women, and for treatment with ACE inhibitors among women. Finally, a decreased stroke mortality was found for treatment with statins among women (HR 0.85, 95% CI 0.77‐0.94).
TABLE 5.
The association between dispensed drugs and the risk of mortality in patients with atrial fibrillation a
| Men | Women | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | |||
| All medication | 1.29 | 1.21 | 1.37 | 1.16 | 1.08 | 1.25 |
| a. Loop diuretics | 1.54 | 1.45 | 1.64 | 1.34 | 1.27 | 1.41 |
| b. Thiazides | 0.99 | 0.91 | 1.08 | 0.97 | 0.91 | 1.04 |
| c. Potassium‐saving agents | 1.34 | 1.18 | 1.53 | 1.08 | 0.99 | 1.19 |
| d. Beta‐1‐selective beta‐blockers | 0.94 | 0.90 | 0.98 | 0.94 | 0.91 | 0.98 |
| e. Non‐selective beta‐blockers | 0.86 | 0.79 | 0.94 | 0.85 | 0.79 | 0.92 |
| f. Heart‐active calcium channel blockers | 1.17 | 1.01 | 1.34 | 0.99 | 0.88 | 1.13 |
| g. Vessel‐active calcium channel blockers | 1.00 | 0.94 | 1.06 | 1.00 | 0.94 | 1.06 |
| h. ACE inhibitors | 1.17 | 1.12 | 1.23 | 1.18 | 1.12 | 1.25 |
| i. ARBs | 0.97 | 0.90 | 1.04 | 0.95 | 0.88 | 1.02 |
| j. Statins | 0.87 | 0.82 | 0.92 | 0.81 | 0.76 | 0.86 |
In the full model, we adjusted for age, region of Sweden, immigrant status, socioeconomic factors, comorbidities, and all drug classes.
Abbreviations: ARBs, angiotensin receptor blocking agents; HR, hazards ratios; CI, confidence intervals.
Full adjusted model.
Bold values are statistically significant.
In further analyses, we analyzed the proportion of dispensed drugs over time (Table S5). The proportion of medication occasions exceeding 5 times were for most drugs 75%, with the lowest rates for potassium‐saving agents (68.2%) and non‐selective beta‐blockers, 69.3%. The rate of multiple medications in the study sample as well as in patients with incident CHF (Tables S6 and S7) was the highest for beta‐1‐selective beta‐blockers, with around 60% for patients with dispensed vessel‐active CCBs, ACE inhibitors, ARBs, and statins. The rates were also high for concomitant loop diuretics, especially in patients diagnosed with incident CHF that had most dispensed agents and with the highest rate among patients with dispensed potassium‐saving agents (79.6%). Otherwise, the rate varied between 35% and 58%. In a competing risk analysis (Table S8), with death as competing risk, the fully adjusted model for beta‐1‐selective beta‐blockers showed low HRs for men (0.95, 95% CI 0.93‐0.98) and women (0.94, 95% CI 0.91‐0.98), as well as among men for non‐selective beta‐blockers (0.90, 95% CI 0.86‐0.95) and statins (0.89, 95% CI 0.86‐0.92). Statistically significant higher HRs were found for loop diuretics, potassium‐saving agents and ACE inhibitors in men and women, and for heart‐active CCBs and ARBs in women.
4. DISCUSSION
The main finding of this study was that both men and women with AF who were treated with beta‐1‐selective beta‐blockers, and men with AF who were treated with non‐selective beta‐blockers, and statins had a lower risk of incident CHF, but also a lower mortality related to heart diseases. On the other hand, some drugs were associated with a higher risk of CHF among both men and women, that is, loop diuretics, potassium‐saving agents, ACE inhibitors and ARBs, and among women also heart‐active CCBs.
The results in the present study differ from an earlier, smaller study from Swedish primary care, 13 where thiazides, vessel‐active CCBs, and non‐selective beta‐blockers were associated with a lower risk of incident CHF in patients with AF. Furthermore, our results differ from studies on patients with hypertension, and as the risk profile is different between patients, with or without AF, finding somewhat different patterns is not surprising.
Beta‐blockers, especially beta‐1‐selective beta‐blockers, but also non‐selective beta‐blockers among men, were associated with a lower risk of incident CHF, but also in general with a lower mortality risk in heart diseases supporting our findings on CHF, as CHF is a great contributor to mortality in AF. 7 Studies on beta‐blockers in hypertension and the risk of CHF have shown conflicting results, with one review finding no positive effect, 12 and another a preventive effect. 16 In the earlier, smaller study involving Swedish primary care patients, non‐selective beta‐blockers (sotalol), were also found to be associated with a lower risk of CHF. 13 Both rate and rhythm regulation of AF could be effective in reducing the risk of CHF in AF, by preventing tachycardia‐mediated cardiomyopathy, 8 and beta‐blockers are valuable in both these strategies.
For statins, we found a lower CHF risk in men, and among women only in the age‐group 65‐84 years. The lower risk was observed also in men and women with an earlier diagnosed heart disease, and among men with CHD. Statins were also associated with a lower mortality risk in heart diseases, and in cardiovascular diseases in general. In the earlier mentioned study for primary care, a lower mortality in general was found in men and women below 80 years of age. 17
The increased risk of CHF in patients treated with RAS‐blockers could seem confusing, as RAS‐blockers have previously been found to prevent CHF in patients with hypertension. 12 , 16 The results in this study could be influenced by different factors, such as by competing risk especially among elderly individuals, and also confounding by indication 18 hence why we also analyzed results in patients with prevalent heart disease or not and found similar results in these subgroups.
Regarding the possible association between diuretics and incident CHF, one review on hypertensive patients found a preventive effect of diuretics in general when used as antihypertensive therapy. 12 However, we found an excess risk of both incident CHF and heart mortality associated with the use of loop diuretics and potassium‐saving agents. One possible explanation is that these patients did have symptoms associated with CHF before they were diagnosed with CHF. However, loop diuretics are also associated in worsening renal function, 19 and also with activation of the renal‐angiotensin‐aldosterone system and sympathetic nervous system, which may be related to our findings. 20 Regarding thiazides, a slightly lower risk for CHF was found among women without heart disease, and a borderline lower risk for women 85 years of age and older.
The associations between CCBs and the risk of incident CHF are also of interest. In patients with hypertension, previous studies have shown conflicting results of CCBs, with two reviews reporting a reduced rate of CHF, 12 , 21 one finding no association, 16 and two reviews finding an increased risk of incident CHF. 22 , 23 As CCBs are associated with edema, this mechanism might increase the risk of incident CHF and might lead to a concomitant prescription of loop diuretics (46.8% for patients with incident CHF and dispensed with vessel‐active CCBs). 24 We found somewhat different patterns in men and women, and in patients with or without heart diseases in general or CHD.
5. LIMITATIONS
There are several limitations with this observational study, which must be kept in mind when interpreting the results, for example, residual confounding. We restricted the assessment period for dispensed drugs between July 1, 2005, and December 31, 2006, and on at least two occasions as a proxy of a persistent drug treatment. However, 75% of medications were dispensed on 5 or more occasions. The findings may have been subject to bias, for example, confounding by indication, 18 or survival bias. To account for this effect, we also analyzed data with or without concomitant CHD, and as a sensitivity analysis also with or without heart diseases in general. We also performed a competing risk analysis, with similar results. We used data from registers, where clinical data, such as values of blood pressure or image diagnostics, for example, echo‐cardiography or coronary angiography, were not at hand. Some diagnoses could be expected to be under‐reported, especially obesity and sleep apnea syndrome. Furthermore, this seems to be true also for hypertension and diabetes, as most of these patients do receive their care in primary care, 25 and we did not have access to the primary care diagnoses. Besides, the number of comorbidities and cardiovascular drugs used seem remarkably low, owing to the methodology we used. Another remark is that we included patients with an AF diagnosis from several years back, but in the sensitivity analysis including patients with a diagnosis during the last 2 years the results were quite similar. One question raised by the findings, especially on the results from loop diuretics, is that the CHF diagnosis seems to be set rather late in the course of the disease, that is, loop diuretics seem to be used before a definite CHF diagnosis is set. For CHF diagnoses registered in hospital records, good validity has been shown, 26 and in a meta‐analysis, the sensitivity of a CHF diagnosis from registers was found to be around 75%. 27 We did not have access to doses of the retrieved drugs. Severity of CHF and CHD was not available. As severity of CHF is an important factor for mortality, this is also a major limitation of the study. Data on ejection fraction and criteria for diagnosis of CHF were not available. Moreover, AF could not be classified as paroxysmal, persistent or permanent, and heart rhythm could not be classified as sinus rhythm or fibrillation rhythm. According to a review, around 50% of AF in Europe is regarded as permanent, and 25% each paroxysmal and persistent, respectively. 2 In addition, we did not have access to kidney function data. All these mentioned factors could have affected our results.
A major strength of this study was that we were able to link data on dispensed drugs to data from national demographic and socioeconomic registers. We used the Swedish Prescribed Drug Register, 14 , 15 in which dispensed prescriptions of drugs are registered and thus showed a higher probability of being used compared with only prescription data.
6. CONCLUSIONS
In conclusion, our findings suggest that the risk of developing CHF may be reduced in AF patients by use of beta‐blockers, and to some extent statins. The use of these drugs in patients with AF is also supported by our results showing a lower risk of mortality in heart diseases. However, we believe that the present observational study needs to be interpreted with caution, as confounding by indication may explain some of our results. More studies focusing on patients with AF with data on severity of CHF, ejection fraction, and kidney function are needed to confirm our findings.
CONFLICT OF INTEREST
The authors have no conflict of interest to disclose.
AUTHOR CONTRIBUTIONS
JS and KS had access to the databases used and provided funding; PW suggested and planned the study, in collaboration with ACC, XL, JS, and KS; XL researched data and performed statistical analyses; PW drafted the manuscript, in collaboration with ACC and edited the manuscript after comments from all co‐authors ACC, XL, MH, JS, and KS edited and critically revised the manuscript and contributed to discussion.
ETHICAL APPROVAL
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The study was approved by the regional ethics boards at Karolinska Institutet and Lund University. Informed consent was not applicable, as the study was based on anonymized data from registers.
Supporting information
Table S1‐S8
ACKNOWLEDGMENTS
We thank Patrick Reilly for language editing.
Wändell P, Carlsson AC, Li X, Holzmann MJ, Sundquist J, Sundquist K. Use of cardiovascular drugs and risk of incident heart failure in patients with atrial fibrillation. J Clin Hypertens. 2020;22:1396–1405. 10.1111/jch.13931
Funding information
This work was supported by ALF funding awarded to Jan Sundquist and Kristina Sundquist and by grants from the Swedish Research Council (awarded to Jan and Kristina Sundquist). Research reported in this publication was also supported by the Swedish Heart‐Lung foundation.
REFERENCES
- 1. GBD 2016 Disease and Injury Incidence and Prevalence Collaborators . Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990‐2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet. 2017;90:1211‐1259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Zoni‐Berisso M, Lercari F, Carazza T, Domenicucci S. Epidemiology of atrial fibrillation: European perspective. Clin Epidemiol. 2014;6:213‐220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Forslund T, Wettermark B, Wandell P, von Euler M, Hasselstrom J, Hjemdahl P. Risk scoring and thromboprophylactic treatment of patients with atrial fibrillation with and without access to primary healthcare data: experience from the Stockholm health care system. Int J Cardiol. 2013;170:208‐214. [DOI] [PubMed] [Google Scholar]
- 4. Friberg L, Bergfeldt L. Atrial fibrillation prevalence revisited. J Intern Med. 2013;274:461‐468. [DOI] [PubMed] [Google Scholar]
- 5. Hobbs FR, Taylor CJ, Jan Geersing G, Rutten FH, Brouwer JR. European Primary Care Cardiovascular Society (EPCCS) consensus guidance on stroke prevention in atrial fibrillation (SPAF) in primary care. Eur J Prev Cardiol. 2016;23(5):460–473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Wann LS, Curtis AB, Ellenbogen KA, et al. Management of patients with atrial fibrillation (compilation of 2006 ACCF/AHA/ESC and 2011 ACCF/AHA/HRS recommendations): a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. Circulation. 2013;127:1916‐1926. [DOI] [PubMed] [Google Scholar]
- 7. O'Neal WT, Salahuddin T, Broughton ST, Soliman EZ. Atrial fibrillation and cardiovascular outcomes in the elderly. Pacing Clin Electrophysiol. 2016;39:907‐913. [DOI] [PubMed] [Google Scholar]
- 8. Seiler J, Stevenson WG. Atrial fibrillation in congestive heart failure. Cardiol Rev. 2010;18:38‐50. [DOI] [PubMed] [Google Scholar]
- 9. Ling LH, Kistler PM, Kalman JM, Schilling RJ, Hunter RJ. Comorbidity of atrial fibrillation and heart failure. Nat Rev Cardiol. 2016;13:131‐147. [DOI] [PubMed] [Google Scholar]
- 10. Wandell P, Carlsson AC, Holzmann MJ, Arnlov J, Sundquist J, Sundquist K. Mortality in patients with atrial fibrillation and common co‐morbidities ‐ a cohort study in primary care. Ann Med. 2018;50:156‐163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wandell P, Carlsson AC, Holzmann MJ, Arnlov J, Sundquist J, Sundquist K. The association between relevant co‐morbidities and prevalent as well as incident heart failure in patients with atrial fibrillation. J Cardiol. 2018;72:26‐32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Sciarretta S, Palano F, Tocci G, Baldini R, Volpe M. Antihypertensive treatment and development of heart failure in hypertension: a Bayesian network meta‐analysis of studies in patients with hypertension and high cardiovascular risk. Arch Intern Med. 2011;171:384‐394. [DOI] [PubMed] [Google Scholar]
- 13. Wandell P, Carlsson AC, Holzmann MJ, Arnlov J, Sundquist J, Sundquist K. Associations between relevant cardiovascular pharmacotherapies and incident heart failure in patients with atrial fibrillation: a cohort study in primary care. J Hypertens. 2018;36:1929‐1935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Wallerstedt SM, Wettermark B, Hoffmann M. The first decade with the Swedish Prescribed Drug Register ‐ a systematic review of the output in the scientific literature. Basic Clin Pharmacol Toxicol. 2016;119:464‐469. [DOI] [PubMed] [Google Scholar]
- 15. Wettermark B, Hammar N, Fored CM, et al. The new Swedish Prescribed Drug Register–opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf. 2007;16:726‐735. [DOI] [PubMed] [Google Scholar]
- 16. Zanchetti A, Thomopoulos C, Parati G. Randomized controlled trials of blood pressure lowering in hypertension: a critical reappraisal. Circ Res. 2015;116:1058‐1073. [DOI] [PubMed] [Google Scholar]
- 17. Wandell P, Carlsson AC, Sundquist K, Johansson SE, Sundquist J. Effect of cardiovascular drug classes on all‐cause mortality among atrial fibrillation patients treated in primary care in Sweden: a cohort study. Eur J Clin Pharmacol. 2013;69:279‐287. [DOI] [PubMed] [Google Scholar]
- 18. Delgado‐Rodriguez M, Llorca J. Bias. J Epidemiol Community Health. 2004;58:635‐641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Carubelli V, Metra M, Lombardi C, et al. Renal dysfunction in acute heart failure: epidemiology, mechanisms and assessment. Heart Fail Rev. 2012;17:271‐282. [DOI] [PubMed] [Google Scholar]
- 20. Chiong JR, Cheung RJ. Loop diuretic therapy in heart failure: the need for solid evidence on a fluid issue. Clin Cardiol. 2010;33:345‐352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Costanzo P, Perrone‐Filardi P, Petretta M, et al. Calcium channel blockers and cardiovascular outcomes: a meta‐analysis of 175,634 patients. J Hypertens. 2009;27:1136‐1151. [DOI] [PubMed] [Google Scholar]
- 22. Shibata MC, Leon H, Chatterley T, Dorgan M, Vandermeer B. Do calcium channel blockers increase the diagnosis of heart failure in patients with hypertension? Am J Cardiol. 2010;106:228‐235. [DOI] [PubMed] [Google Scholar]
- 23. Sica DA, Carter B, Cushman W, Hamm L. Thiazide and loop diuretics. J Clin Hypertens (Greenwich). 2011;13:639‐643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Woodford HJ. Calcium channel blockers co‐prescribed with loop diuretics: a potential marker of poor prescribing? Drugs Aging. 2020;37:77‐81. [DOI] [PubMed] [Google Scholar]
- 25. Carlsson AC, Wandell P, Osby U, Zarrinkoub R, Wettermark B, Ljunggren G. High prevalence of diagnosis of diabetes, depression, anxiety, hypertension, asthma and COPD in the total population of Stockholm, Sweden ‐ a challenge for public health. BMC Public Health. 2013;13:670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Ingelsson E, Arnlov J, Sundstrom J, Lind L. The validity of a diagnosis of heart failure in a hospital discharge register. Eur J Heart Fail. 2005;7:787‐791. [DOI] [PubMed] [Google Scholar]
- 27. McCormick N, Lacaille D, Bhole V, Avina‐Zubieta JA. Validity of heart failure diagnoses in administrative databases: a systematic review and meta‐analysis. PLoS One. 2014;9:e104519. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1‐S8