Abstract

Objective:

To examine associations between early anxiety–depression symptoms, perceived social support, and in-hospital outcomes in patients with acute myocardial infarction (AMI).

Methods:

This observational cross-sectional study enrolled 150 consecutive AMI patients. Anxiety–depression symptoms and perceived social support were assessed within 24–72 hours of admission using the Hospital Anxiety and Depression Scale (HADS) and the Perceived Social Support Scale (PSSS). Primary outcome was in-hospital complications; secondary outcomes were length of stay and sleep quality. Multivariable regression models were applied with adjustment for age, cardiac function, emergency PCI, and major comorbidities.

Results:

Clinically significant anxiety–depression symptoms (HADS ≥11) were present in 44.0% of patients; 27.3% developed at least one complication. Higher HADS scores were independently associated with increased complication risk, prolonged stay, and poorer sleep quality (all P < 0.05). Higher PSSS scores were associated with reduced complication risk, shorter stay, and better sleep quality (all P < 0.05). Psychosocial risk stratification demonstrated a significant gradient across all outcomes (trend P < 0.05), with results consistent across subgroup and sensitivity analyses.

Conclusion:

Early anxiety–depression symptoms and perceived social support are independently associated with in-hospital outcomes in AMI patients. Routine psychosocial screening may help identify high-risk patients and guide individualized nursing care.

1 Introduction

Acute myocardial infarction (AMI) is a severe cardiovascular condition characterized by sudden onset, rapid progression, and high risks of mortality and disability. Although reperfusion strategies, particularly emergency percutaneous coronary intervention (PCI), and evidence-based pharmacological therapy have markedly improved survival in patients with AMI, complications during hospitalization, prolonged hospital stay, and poor inpatient experience remain common. These issues adversely affect short-term recovery and healthcare resource utilization (1). Therefore, in addition to disease severity and treatment strategies, identifying other factors associated with in-hospital outcomes in patients with AMI is of clinical and nursing importance.

Beyond traditional clinical and physiological factors, psychosocial factors have received increasing attention in cardiovascular disease onset, progression, and prognosis (2). AMI is a major acute stress event and is frequently accompanied by psychological reactions. Symptoms of anxiety and depression are common among hospitalized patients (3, 4). Previous studies have shown that anxiety–depression symptom burden affects subjective well-being and quality of life. It may also impair cardiovascular stability through neuroendocrine imbalance, enhanced inflammatory responses, and behavioral changes (5, 6). However, most existing studies focus on medium- or long-term prognosis after AMI. The independent role of psychological factors in short-term in-hospital outcomes remains insufficiently evaluated (7, 8). Importantly, psychological symptoms assessed during the early stage of hospitalization may reflect the acute stress response to myocardial infarction and may directly influence treatment adherence, autonomic function, inflammatory activity, and sleep quality, thereby affecting short-term clinical outcomes (5, 9). Compared with assessments conducted during the recovery or follow-up period, early assessment provides a time-sensitive opportunity to identify high-risk patients and implement timely interventions (10). However, evidence regarding the clinical relevance of early anxiety and depression symptoms during the acute hospitalization phase remains limited.

Social support is an important psychosocial resource and is considered to have buffering and protective effects during acute illness and recovery (11). Adequate family and social support can reduce psychological stress, improve emotional status, and enhance understanding and adherence to treatment and nursing care (12). Lack of family support has also been linked to increased long-term mortality in patients with AMI (13, 14). However, studies that systematically assess social support during the acute hospitalization phase and examine its association with in-hospital outcomes are still limited (15). In addition, anxiety–depression symptoms and social support may interact with each other. Assessment from a single dimension may not fully capture psychosocial risk (16). In clinical practice, inpatient risk assessment in AMI mainly relies on disease severity and treatment response, and the role of psychosocial factors in early evaluation remains insufficiently addressed (17). Therefore, it is necessary to investigate the associations between early anxiety–depression symptom burden, social support, and in-hospital outcomes in real-world inpatient settings and to explore psychosocial risk stratification for inpatient management.

In this study, “early anxiety–depression” does not refer to previously diagnosed psychiatric disorders, but rather to the symptom burden of anxiety and depression experienced by patients during the acute phase of AMI hospitalization (within 24–72 hours after admission), as measured using the Hospital Anxiety and Depression Scale (HADS). This concept is distinct from chronic or pre-existing anxiety and depressive disorders and focuses on capturing the immediate psychological response triggered by an acute cardiovascular event.

2 Materials and methods2.1 Study design and participants

This study employed an observational cross-sectional design. Psychosocial assessments were conducted once within 24–72 hours after admission, and in-hospital outcomes were recorded at discharge. No follow-up beyond hospitalization was performed, and all regression analyses were based on this data structure. Patients with AMI who were hospitalized in the Department of Cardiology of The First Affiliated Hospital of Soochow University between October 1, 2024 and October 1, 2025 were consecutively enrolled. A total of 150 eligible patients were included. AMI was diagnosed according to current clinical guidelines based on symptoms, electrocardiographic changes, and elevated cardiac biomarkers. Inclusion criteria were: (1) age ≥ 18 years; (2) confirmed diagnosis of AMI; (3) completion of HADS and social support assessment within 24–72 hours after admission; and (4) complete clinical and outcome data during hospitalization. Exclusion criteria were: (1) prior diagnosis of severe psychiatric disorders or cognitive impairment; (2) inability to complete psychological assessment due to critical illness or communication barriers; and (3) missing key variables or outcome data. The study protocol was approved by the Ethics Committee of The First Affiliated Hospital of Soochow University. The study used routinely documented clinical data, nursing assessment records, and psychosocial assessment data recorded as part of standard inpatient care. Access to identifiable records was restricted to authorized investigators under institutional ethical approval, and all data were de-identified prior to analysis to ensure confidentiality protection. As the study involved minimal risk to participants, no additional patient contact, and no intervention beyond routine care, the Ethics Committee waived the requirement for written informed consent. The process of patient identification, eligibility assessment, exclusion, and final inclusion is summarized in Figure 1.

Flow diagram of patient selection and analysis.

2.2 Data collection2.2.1 General clinical data

Demographic and clinical data were collected, including age, sex, marital status, educational level, type of myocardial infarction (ST-segment elevation myocardial infarction [STEMI] or non-ST-segment elevation myocardial infarction [NSTEMI]), receipt of emergency PCI, and cardiac function (Killip classification, I–IV). Age was categorized into three groups (<60, 60–69, and ≥70 years) based on commonly used clinical stratification. Major comorbidities, including hypertension, diabetes mellitus, dyslipidemia, and chronic kidney disease, were recorded and identified based on documented clinical diagnoses in medical records. In addition, lifestyle-related factors, including smoking status (current smoker vs. non-current smoker), were collected. All variables were extracted from electronic medical records and nursing assessment records.

2.2.2 Psychosocial measures

Anxiety and depression symptom assessment. Symptoms of anxiety and depression were assessed using the HADS (18). The scale contains 14 items, including the anxiety (HADS-A) and depression (HADS-D) subscales, with seven items each. Each item is scored from 0 to 3, with a total score range of 0–42. Higher scores indicate more severe symptoms. HADS scores obtained within 24–72 hours after admission were used as baseline values. A HADS total score ≥11 was considered indicative of clinically significant anxiety–depression symptoms and was used to define a high psychological symptom burden in this study (18). If multiple records were available, the earliest complete assessment was selected. In the primary analysis, the HADS total score was treated as a continuous variable to evaluate the association between each one-point increase and in-hospital outcomes.

Social support assessment. Perceived social support was measured using the Chinese version of the Perceived Social Support Scale (PSSS), which was originally developed by Zimet et al. and translated into Chinese by Jiang (19–21). The scale includes 12 items covering family, friend, and other support. Each item is rated from 1 to 7, yielding a total score range of 12–84. Higher scores indicate higher perceived support. PSSS scores completed within 24–72 hours after admission were used as baseline measures. The total score was entered into regression models as a continuous variable.

Variable categorization in sensitivity analysis. HADS scores were dichotomized as high risk (≥11) or low risk (<11) based on established scoring criteria from the original scale development (18). PSSS scores were categorized as low (≤36) or moderate-to-high (>36) social support, consistent with the standard scoring categories of the Chinese version (19–21). To reduce dependence on specific cut-offs, quartile-based stratification was also applied (22). Patients were divided into four groups (Q1–Q4) according to score distribution. The lowest-risk groups (HADS Q1; PSSS Q4) were used as reference categories.

Psychosocial risk stratification. A psychosocial risk stratification index was constructed based on early anxiety–depression symptoms and social support levels (2). Patients were classified into low-risk (low HADS and high PSSS), intermediate-risk (low HADS and low PSSS, or high HADS and high PSSS), and high-risk (high HADS and low PSSS) groups. This stratification was used for comparative outcome analyses.

2.2.3 Outcome measures

The primary outcome was the occurrence of in-hospital complications. Complications were recorded based on predefined criteria and coded as present or absent. Events included severe arrhythmia, worsening heart failure, cardiogenic shock, and transfer to the intensive care unit.

Secondary outcomes were length of hospital stay and sleep quality. Sleep quality was assessed using Pittsburgh Sleep Quality Index (PSQI) scores routinely recorded in the nursing assessment system during hospitalization (23). Based on the sample distribution, the 33rd percentile (P33) and 67th percentile (P67) of PSQI scores were used to classify patients into good (≤P33), moderate (>P33 to ≤P67), and poor (>P67) sleep quality groups. Length of stay was defined as the number of days from admission to discharge and analyzed as a continuous variable.

2.3 Statistical analysis

Statistical analyses were performed using SPSS. Continuous variables were expressed as mean ± standard deviation (mean ± SD) or median (interquartile range, IQR). Categorical variables were presented as number and percentage [n (%)]. Group comparisons used the t test or Mann–Whitney U test for continuous variables and the chi-square test for categorical variables. Multivariable logistic regression was applied to in-hospital complications. Ordinal logistic regression was used for sleep quality, and multiple linear regression was used for length of stay. Potential confounders were adjusted according to study design and outcome characteristics. Basic models included age, sex, and major clinical features. Emergency PCI, Killip class, major comorbidities, and psychosocial factors were added sequentially. Trend analyses for psychosocial risk stratification were conducted using the Cochran–Armitage test or the Jonckheere–Terpstra test. Subgroup and sensitivity analyses were performed. All tests were two-sided, and P < 0.05 was considered statistically significant.

A formal a priori sample size calculation was not performed, as the study consecutively enrolled all eligible patients during the study period. To address this limitation, a post-hoc power analysis was conducted for the primary logistic regression model. The unadjusted odds ratio for clinically significant anxiety–depression symptoms (HADS ≥11 vs. <11; OR = 2.59) was derived from univariate analysis and used as the effect size estimate. With a baseline complication rate of 27.3%, two-sided α = 0.05, and 7 predictors in the final model, the minimum sample size for 80% power was estimated at approximately 75 participants. The present study enrolled 150 participants, exceeding this threshold twofold and indicating adequate statistical power. The power calculation was performed using G*Power (version 3.1.9.6). Of note, HADS and PSSS were entered as continuous variables in the primary models, yielding inherently modest per-unit odds ratios. The power estimate was therefore based on a binary categorical comparison using the prespecified cut-off, providing a more interpretable effect size reference. In multivariable regression analyses, in-hospital complications were defined as the primary outcome, with a total of 41 events observed. Given the relatively limited number of events, the number of variables included in the final models was restricted. Only clinically relevant variables that were associated with the outcome in univariate analyses were retained to reduce the risk of overfitting and ensure model stability.

3 Results3.1 General clinical characteristics

General clinical characteristics of the 150 enrolled patients are presented in Table 1. Psychosocial factors and in-hospital outcomes are summarized in Table 2. Of note, 44.0% of patients had clinically significant anxiety–depression symptoms, 32.7% had low perceived social support, and 27.3% developed at least one in-hospital complication.

VariableCategoryn%Age (years)<605536.760–695234.7≥704328.6SexMale11274.7Female3825.3Marital statusMarried12885.3Unmarried/Divorced/Widowed2214.7Education levelJunior high school or below5436.0High school/Technical secondary school6342.0Junior college or above3322.0MI typeSTEMI9664.0NSTEMI5436.0Emergency PCIYes10469.3No4630.7Killip classI9261.3II3624.0III1510.0IV74.7ComorbiditiesHypertension8858.7Diabetes5436.0Dyslipidemia7248.0Chronic kidney disease149.3Current smokingYes7952.7No7147.3

Demographic and clinical characteristics of patients hospitalized with acute myocardial infarction (n = 150).

STEMI, ST-segment elevation myocardial infarction; NSTEMI, non–ST-segment elevation myocardial infarction.

VariableValueHADS total score (0–42), mean ± SD15.2 ± 7.4HADS-A (anxiety, 0–21), mean ± SD8.1 ± 4.1HADS-D (depression, 0–21), mean ± SD7.1 ± 4.3HADS categories, n (%)Normal (0–7): 46 (30.7%)Borderline (8–10): 38 (25.3%)Definite (≥11): 66 (44.0%)PSSS total score (12–84), n (%)56.8 ± 12.6PSSS categories, n (%)Low (≤50): 49 (32.7%)Moderate (51–62): 51 (34.0%)High (≥63): 50 (33.3%)Any in-hospital complication, n (%)41 (27.3%)Complication components, n (%)Severe arrhythmia: 18 (12.0%)worsening heart failure: 15 (10.0%)cardiogenic shock: 9 (6.0%)ICU transfer: 11 (7.3%)Length of stay (days), median (IQR)8 (6–11)Length of stay categories, n (%)≤7 days: 56 (37.3%)8–10 days: 51 (34.0%)≥11 days: 43 (28.7%)Sleep quality, n (%)Good: 59 (39.3%)Fair: 63 (42.0%)Poor: 28 (18.7%)

Descriptive statistics of psychosocial factors and in-hospital outcomes (n = 150).

“Any in-hospital complication” was defined as the occurrence of at least one complication during hospitalization (n = 41). The proportions of individual complications were calculated using the total sample size (n = 150) as the denominator. As multiple complications may occur in the same patient, the sum of individual components may exceed 41.

3.2 Univariate analysis by in-hospital complication status

Univariate analyses indicated that age group, Killip class, chronic kidney disease, anxiety–depression level, and social support level were significantly associated with the occurrence of in-hospital complications (P < 0.05) (Table 3).

VariableCategoryNo complications (n = 109)Complications (n = 41)StatisticP valueAge (years), n (%)<6045 (41.3)10 (24.4)60–6939 (35.8)13 (31.7)χ2 = 8.210.016≥7025 (22.9)18 (43.9)Sex, n (%)Male80 (73.4)32 (78.0)χ2 = 0.330.565Female29 (26.6)9 (22.0)Marital status, n (%)Married95 (87.2)33 (80.5)χ2 = 1.070.301Unmarried/Divorced/Widowed14 (12.8)8 (19.5)Education level, n (%)Junior high or below38 (34.9)16 (39.0)High school/Technical secondary45 (41.3)18 (43.9)χ2 = 0.860.651Junior college or above26 (23.9)7 (17.1)MI type, n (%)STEMI66 (60.6)30 (73.2)χ2 = 2.140.143NSTEMI43 (39.4)11 (26.8)Emergency PCI, n (%)Yes80 (73.4)24 (58.5)χ2 = 3.270.070No29 (26.6)17 (41.5)Killip class, n (%)I–II98 (89.9)30 (73.2)χ2 = 7.120.008III–IV11 (10.1)11 (26.8)Hypertension, n (%)No48 (44.0)14 (34.1)Yes61 (56.0)27 (65.9)χ2 = 1.180.277Diabetes, n (%)No73 (67.0)23 (56.1)Yes36 (33.0)18 (43.9)χ2 = 1.610.205Dyslipidemia, n (%)No61 (56.0)17 (41.5)Yes48 (44.0)24 (58.5)χ2 = 2.630.105Chronic kidney disease, n (%)No102 (93.6)34 (82.9)Yes7 (6.4)7 (17.1)χ2 = 4.090.043Current smoking, n (%)No53 (48.6)18 (43.9)Yes56 (51.4)23 (56.1)χ2 = 0.260.609HADS category, n (%)<840 (36.7)6 (14.6)8–1028 (25.7)10 (24.4)χ2 = 10.60.005≥1141 (37.6)25 (61.0)PSSS category, n (%)High (≥63)45 (41.3)5 (12.2)Moderate (51–62)39 (35.8)12 (29.3)χ2 = 18.1<0.001Low (≤50)25 (22.9)24 (58.5)Length of stay, days, median (IQR)—7 (6–10)10 (8–13)U = 1390<0.001Sleep quality, n (%)Good/Fair95 (87.2)27 (65.9)Poor14 (12.8)14 (34.1)χ2 = 9.490.002

Univariate comparisons between patients with and without in-hospital complications (n = 150).

In-hospital complications were a composite outcome, defined as the occurrence of any of the following events: severe arrhythmia, worsening heart failure, cardiogenic shock, or ICU transfer. Categorical variables were compared using the chi-square test. Continuous variables were compared using the Mann–Whitney U test according to distribution. Abbreviations: STEMI, ST-segment elevation myocardial infarction; NSTEMI, non–ST-segment elevation myocardial infarction.

3.3 Multivariable logistic regression for in-hospital complications

Based on the univariate results, variables with P < 0.10 were included in multivariable logistic regression models. The dependent variable was the occurrence of in-hospital complications (yes/no). The analysis aimed to evaluate the independent associations between psychosocial factors and complications. Multivariable logistic regression analysis is presented in Table 4. After adjustment for clinical covariates, higher HADS total scores were independently associated with an increased risk of in-hospital complications, whereas higher PSSS total scores were independently associated with a reduced risk.

VariableModel 1 (OR, 95% CI)P valueModel 2 (OR, 95% CI)P valueAge 60–69 years1.42 (0.63–3.18)0.3921.31 (0.57–3.03)0.523Age ≥70 years2.56 (1.14–5.74)0.0232.21 (1.01–5.13)0.046Emergency PCI (yes)0.62 (0.31–1.03)0.0660.69 (0.35–1.17)0.128Killip class III–IV2.98 (1.28–6.94)0.0112.41 (1.02–5.72)0.045Chronic kidney disease2.63 (1.01–6.83)0.0482.19 (0.82–5.87)0.119HADS total score (per 1-point increase)——1.08 (1.03–1.14)0.002PSSS total score (per 1-point increase)——0.96 (0.93–0.99)0.008

Multivariable logistic regression analysis for in-hospital complications (n = 150).

Model 1 adjusted for age group, emergency PCI, Killip class, and chronic kidney disease. Model 2 further included HADS total score and PSSS total score. The same adjustment principle applied below.

3.4 Associations between psychosocial factors and length of hospital stay

Length of stay was treated as a continuous outcome. Multiple regression models were constructed to examine the associations between psychosocial factors and the hospitalization course. Variables with P < 0.10 in univariate analyses for length of stay were included. (Supplementary Table 1).

In the multivariable regression model, higher HADS total scores were independently associated with longer hospital stay, whereas higher PSSS total scores were independently associated with shorter hospitalization (both P < 0.05) (Table 5).

VariableModel 1 (β, 95% CI)P valueModel 2 (β, 95% CI)P valueAge 60–69 years0.72 (−0.48, 1.92)0.2380.61 (−0.57, 1.79)0.312Age ≥70 years1.84 (0.63, 3.05)0.0031.52 (0.36, 2.68)0.010Emergency PCI (yes)−0.96 (−2.01, 0.09)0.073