Associations Between Physical and Affective Symptom Clusters and 6-Month Outcomes in HF

Study Questions:

How are independent physical and affective symptom clusters in heart failure (HF) related, and does this knowledge illuminate important factors influencing clinical outcomes that might otherwise remain obscured?


This secondary analysis of data from two parallel prospective cohort studies of symptoms among adults with moderate to advanced HF (New York Heart Association Class II-IV) conducted at the same institution in the Pacific Northwest included in the data modelling analysis demographic and clinical characteristics data, common physical symptoms (dyspnea and sleep-wake disturbances) and affective symptoms (depression and anxiety), and 6-month clinical outcome events. The Heart Failure Somatic Perception Scale Dyspnea Subscale and Epworth Sleepiness Scale were used to measure dyspnea and awake-sleep disturbances, respectively, and the Patient Health Questionnaire-9 and Brief Symptom Inventory Anxiety Subscale were used to assess depression and anxiety, respectively. Electronic medical records were used to extract outcome event data, as were follow-up telephone interviews with participants.


The combined study sample (n = 274) comprised data for 167 (61%) men; the average age was 57.2 ± 13.2 years. Over half (157 [57.3%]) the sample had New York Heart Association Class III HF, and nearly two-thirds (174 [63.7%]) had nonischemic HF. Physical symptom clusters (dyspnea and awake-sleep disturbance) and affective symptom clusters (depression and anxiety) were grouped according to whether they were mild or severe. Significant differences were found between dyspnea and awake-sleep disturbance scores between the mild and severe physical symptom clusters, and there were significant differences depression and anxiety scores between the two affective symptom clusters. Cross-classification modeling indicated acceptable model fit in symptom clusters associations (entropy = 0.80). Those with mild physical and mild affective symptom clusters were labeled congruent-mild symptoms group (n = 190; 69.3%), and those with severe physical and affective symptom clusters were labeled congruent-severe symptom group (n = 46; 16.8%). A combination symptom cluster group (labeled incongruent symptom group) comprised the small number of participants who had mild physical/severe affective symptoms and severe physical/mild affective symptom clusters (n = 38; 13.9%). Within 6 months, 86 participants had a clinical event. There were 5 deaths, 66 independent hospital admissions for cardiovascular reasons, and 15 emergency department admissions for cardiovascular causes. Five participants (1.8%) were lost to follow-up. Adjusting for the Seattle Heart Failure Model risk score, Cox proportional hazards regression showed those in the incongruent symptom group were 98% more likely to have a clinical event (death, hospital admission, or emergency department admission) within 6 months (p = 0.014) than those in the congruent-mild symptom group (model likelihood ratio χ2 = 20.95; p = 0.0001; Harrell’s C-statistic = 0.65). No statistical differences in clinical events were found between those in the congruent-severe symptom group and those in the congruent-mild symptom group. Few significant differences in sociodemographic and clinical characteristics existed between the symptom subgroups. Using logistic regression with the congruent-mild symptom group as the referent group, a significant determinant of the incongruent symptom group membership was lack of diuretic use, and significant determinants of congruent-severe symptom group membership were aldosterone antagonist use and antidepressant/anxiolytic use.


In this sample, there was a strong association between physical and affective symptom clusters by symptom severity. Participants who had the mild physical symptom cluster were more likely to have the mild affective symptom cluster. Similarly, those who had the severe physical symptom cluster were more likely to have the severe affective symptom cluster. Interestingly, a small group of participants who had incongruent physical and affective symptom clusters (one mild and one severe symptom cluster) had significantly worse 6-month event-free survival.


Symptom burden is high in persons living with HF. This secondary analysis study provides insight on how independent physical and affective symptom clusters align based on symptom severity, and how naturally occurring patterns of alignment predict outcomes. Having discordant severity of physical and affective symptoms has clinical relevance in that patients with incongruent symptom clusters had worse clinical outcomes than those in the congruent symptom cluster groups.

Clinical Topics: Heart Failure and Cardiomyopathies, Acute Heart Failure, Sleep Apnea

Keywords: Heart Failure, Affective Symptoms, Signs and Symptoms, Sleep Stages, Dyspnea, Depression, Anxiety, Electronic Health Records

< Back to Listings