Well-Being Profiles of Family Caregivers of Patients With Dementia From Romania: A Latent Profile Analysis.

INTRODUCTION

The rising prevalence of dementia has increased the demand for long-term care, with family members often assuming caregiving responsibilities. While this form of care reduces healthcare costs and improves patients' quality of life, it also exposes caregivers to physical and mental health challenges, often rendering them "invisible patients." Well-being remains a key focus in both medical and psychosocial research and can be assessed through Ryff's eudaimonic framework. This study aims to identify distinct latent profiles of family caregivers based on well-being patterns, highlighting, through a person-centered approach, the vulnerabilities and resources associated with each profile.

METHODS

The study included 73 family caregivers from Romania, aged between 30 and 87 years (M = 57.12, SD = 10.36), the majority being women (75.3%). Latent profile analysis (LPA) was used to identify well-being patterns based on scores obtained on the six dimensions of Ryff's scale (54 items, adapted for Romania). The selection of models and distinctiveness of profiles were statistically established through various criteria (Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and entropy), together with a meaningful interpretation, emphasizing the role of meaning in life, autonomy, and self-acceptance in maintaining caregivers' health, well-being, and resilience.

RESULTS

The LPA analysis identified a four-profile model of well-being, i.e., high (11%), moderate (38%), low (41.1%), and very low (10%), highlighting both the heterogeneity of perceptions and the caregivers' vulnerabilities and strengths within each profile. Major differences between profiles are primarily driven by the purpose in life and autonomy dimensions, with values progressively decreasing from one profile to the next. The high profile exhibits the highest scores on these dimensions, while the very low profile records the lowest. The most pronounced deficits appear in the very low profile, particularly in self-acceptance and environmental mastery. The high entropy value of the model (0.93) indicates a well-defined solution with significant differences between profiles.

CONCLUSION

The study highlights variations in well-being among family caregivers of individuals with dementia, making a significant contribution to the identification of distinct latent profiles. A person-centered approach facilitates tailored interventions by clinicians, while the findings provide valuable support both for clinical practitioners and for the development of public health policies.