Comment on “Can psychological interventions affect cardiac rehabilitation patients' well-being? Preliminary results from a longitudinal study”

Dear Editor,

We read with interest the study by Zambetti et al. [1], examining the effects of psychoeducational group therapy, progressive muscle relaxation training, and individual counseling on anxiety, depression, and illness perception in 181 cardiac rehabilitation patients assessed at admission and discharge from the inpatient unit.

1. Study design and the non-random group assignment problem

The study addresses a clinically important question in an ecologically valid real-world rehabilitation setting, and the longitudinal design with two measurement points and a relatively low dropout rate of 16.6% is a genuine strength. The general linear model approach incorporating sociodemographic and clinical covariates to account for confounding is appropriate for the research question, and the use of both the Hospital Anxiety and Depression Scale and the Brief Illness Perception Questionnaire provides a reasonably comprehensive picture of patients' psychological functioning across cognitive, emotional, and symptomatic domains. The finding that individual counseling predicted significant reductions in both anxiety and depression, alongside improved illness perception, is clinically meaningful and consistent with the broader literature on targeted psychological support in cardiac populations [2].

The central methodological problem, which the authors acknowledge but do not fully work through in terms of its interpretive implications, is the non-random assignment to the intervention groups. Patients were assigned to progressive muscle relaxation training if they had borderline or pathological HADS scores of 8 or above, and to individual counseling if they had pathological scores of 11 or more. The psychoeducational group was attended by all patients. This allocation structure means that the three intervention groups were not comparable at baseline [3], and patients receiving individual counseling started with systematically higher anxiety and depression levels than those in the relaxation or psychoeducation-only conditions by design. The general linear model accounts for baseline levels as covariates, and the results show the expected pattern: individual counseling patients start higher and end lower. However, this raises a regression-to-the-mean concern that this study does not directly address [4]. Patients with pathological scores at admission are statistically likely to improve regardless of intervention because extreme values tend to move toward the population mean over time [5] and the most severe presentations are most susceptible to natural disease course fluctuation within a two-week inpatient stay that also includes pharmacological optimization, dietary education, and structured physical exercise. Without a control group of similarly distressed patients who received no individual counseling [6], which the design makes impossible by construction, since patients above the threshold were all offered the intervention, the observed reductions attributed to counseling cannot be distinguished from expected natural remission, regression-to-the-mean, or the non-specific effects of a supportive structured inpatient environment where patients have daily contact with a multidisciplinary clinical team. The paper's title and conclusions frame individual counseling as the most high-impact intervention, but this framing is not supported by the study's design. A more defensible and still clinically useful conclusion would be that individual counseling, offered to the most severely distressed patients, was associated with the greatest observed symptom reduction over a two-week inpatient period, which does not establish efficacy over and above what would be expected from time, natural recovery, and intensive multidisciplinary care alone.

2. Statistical validity and the physical comorbidity finding

The finding that patients without physical comorbidity had higher anxiety at T0, while patients with physical comorbidity had higher anxiety at T1, is one of the more clinically interesting results in the study and is explained primarily through appeal to the cardiorenal syndrome and multimorbidity literature. The directional reversal between groups across time points is plausible: patients newly diagnosed with a cardiac condition without prior chronic illness may experience acute distress at admission that resolves with reassurance and rehabilitation, whereas patients managing multiple chronic conditions face an ongoing physiological and psychological burden [7] that may become more salient as the structured support of in-patient care ends. However, the study did not report whether the comorbidity-by-time interaction term itself reached statistical significance [8] or whether the group difference at T1 was directly significant when examined independently. Reporting only that physical comorbidity predicted anxiety reduction without characterizing the nature and statistical robustness of the interaction, specifically whether both groups improved over time and one improved more, or whether one group's anxiety actually increased, makes the finding difficult to translate into clinical guidance. These distinctions are worth clarifying before drawing practical conclusions about how physical comorbidity should modify psychological intervention planning in cardiac rehabilitation settings.

2.1. Clinical applicability and translational context

These methodological observations do not reduce the study's contribution as a carefully conducted, ecologically valid description of psychological outcomes across three concurrent intervention modalities in a functioning cardiac rehabilitation unit. Gathering this kind of real-world longitudinal data in a multidisciplinary inpatient setting is genuinely difficult, and the authors' commitment to following patients to a three-month community-based T2 assessment is the most important next step for this dataset. Whether the improvements observed at discharge, particularly the anxiety and depression reductions in the individual counseling group, are maintained after patients return to community settings, where the daily structure, clinical contact, and social support of inpatient rehabilitation are removed, will be the key question for evaluating the true clinical durability of these interventions. This analysis, combined with a more conservative framing of the T0-to-T1 intervention effects that explicitly acknowledges baseline severity differences and regression-to-the-mean as alternative explanations, would substantially strengthen the conclusions that can be drawn from this dataset and would provide a more reliable evidence base for decisions about psychological staffing and intervention prioritization in cardiac rehabilitation programs.

Ethical approval

Not applicable (commentary on previously published research).

Generative AI disclosure

Generative AI tools were used solely for language refinement and formatting assistance. All scientific interpretation, critique, methodological evaluation, and conceptual analysis were independently developed by the authors.

Funding

None.

CRediT authorship contribution statement

Mrityunjoy Saha: Conceptualization, Data curation, Formal analysis. K. Sharat Vishwanath: Visualization, Writing – original draft, Writing – review & editing.

Conflicts of interest

The authors declare no conflicts of interest.

Data availability

No new data were generated or analysed in this work.