Latent profile analysis of home-based medication self-management behaviours and associated factors in older adults with diabetes.

AIM AND OBJECTIVES: To identify latent profiles of home-based medication self-management behaviours among older adults with diabetes, and to explore their associated factors. BACKGROUND: Suboptimal medication adherence is common among older adults with diabetes. Person-centred behavioural profiles can guide targeted nursing interventions. DESIGN: A cross-sectional study reported in accordance with STROBE. METHODS: We recruited 302 older adults with type 2 diabetes. Data were collected using the Self-Administration of Medication (SAM) scale and a Medication Safety Knowledge and Attitude questionnaire. Latent profile analysis and multinomial logistic regression identified profiles and associated factors. RESULTS: The mean SAM score was 59.93 ± 17.58. Three profiles emerged: low management with low cognitive ability (25.5%), moderate management with high experience (52.6%), and high management with high ability (21.9%). Using the high-management group as reference, retirement (OR = 0.118) and higher knowledge and attitude scores (OR = 0.642; 0.792) reduced the odds of low-management membership, whereas income ≤ 1000 CNY increased the odds (OR = 7.668). Information-seeking (OR = 0.377) and higher attitude scores (OR = 0.875) reduced the odds of moderate-management membership, while irregular follow-ups increased the odds (OR = 5.238). CONCLUSION: Medication self-management in older adults is heterogeneous and suboptimal. Tailored, profile-based nursing interventions are essential to improve adherence and diabetes outcomes. RELEVANCE TO CLINICAL PRACTICE: Recognising these profiles enables nurses to deliver stratified, patient-centred care. Targeted education and support can enhance adherence, reduce risks, and optimise home-based diabetes management.