Background: Symptom trajectories in people with heart failure (HF) can vary over time. Few studies have explored how symptom clusters transition, essential to determine interventions. Aims: (1) Describe transitions in symptom cluster profiles in people with HF over 8 years; (2) Identify factors predicting the transitions. Methods: We conducted secondary analysis of data from the 2008, 2012, and 2016 surveys of the Health and Retirement Study (HRS), using latent transition analysis to identify cluster profiles (fatigue, dyspnea, pain, swelling, depressive symptoms, dizziness), and multinomial regression to determine the predictive factors. Results: HRS included 690 HF subjects (57% female; mean age=74.9 (SD =10.04) years). Four cluster profiles were identified: high burden, low burden, distressing, and respiratory-depressive distress. The probability staying in the same profile ranged from 0.60 to 0.89. Male sex (OR=1.39; 95% CI=0.92, 1.82), black race (OR=1.92; 95% CI=1.37, 2.73), smoking (OR=2.44; 95% CI=1.56, 3.71), having diabetes (OR=3.99; 95% CI=2.90, 6.27) and lung disease (OR=1.40; 95% CI=0.83, 1.68) were associated with an increased odds of moving from the distressing to the high burden profile. Smoking (OR=1.17; 95% CI=0.66, 1.55), having arthritis (OR=4.44; 95% CI=2.80, 6.33), lung disease (OR=1.72; 95% CI=1.17, 2.49), and diabetes (OR=3.84; 95% CI=2.87, 5.94) were associated with an increased odds of moving from the respiratory-depressive distress to the high burden profile. Black race (OR=3.58; 95% CI=2.80, 6.20), smoking (OR=1.44; 95% CI=0.91, 2.10), having arthritis (OR=1.71; 95% CI= 1.20, 2.59), lung disease (OR=1.30; 95% CI= 0.55, 1.95), and diabetes (OR=1.81; 95% CI=1.34, 2.90) were associated with an increased odds of moving from the low burden to the distressing profile. Male sex (OR=2.13; 95% CI= 1.66, 3.68), having arthritis (OR=1.51; 95% CI=1.15, 2.13) and lung disease (OR=2.28; 95% CI=1.57, 4.34) were associated with an increased odds of moving from the low burden to the respiratory-depressive distress profile. Conclusions: Although symptom cluster profiles remained stable, transitions were observed, and deterioration can occur. The identified predictive factors may enhance long-term symptom management.
Read full abstract