Latent Profile Analysis of Sleep Subtypes in Older Adults with Subjective Cognitive Decline and Its Influencing Factors

Abstract

Background Sleep disorders combined with subjective cognitive decline (SCD) in older adults are associated with an increased risk of cognitive decline and dementia conversion. However, sleep problems in older adults with SCD have not received sufficient attention, the sleep subtypes of older adults with SCD and their influencing factors need to be further investigated. Objective To explore potential sleep subtypes in older adults with SCD and analyze the influencing factors of different sleep subtypes. Methods From May to August 2022, older adults with SCD were selected as subjects from the communities in Nanjing, Changzhou, Nantong, and Xuzhou in Jiangsu Province using a stratified convenience sampling method. The general information questionnaire, Subjective Cognitive Decline Questionnaire (SCD-Q9), Beijing Version of the Montreal Cognitive Assessment (MoCA), Mini-Mental State Examination (MMSE), Pittsburgh Sleep Quality Index (PSQI), Patient Health Questionnaire-9 (PHQ-9) and Fatigue, Resistance, Ambulation, Illness and Loss of Weight Index (FRAIL) were used to conduct the survey. The latent profile analysis of sleep in older adults with SCD was performed based on the dimension scores of the PSQI scale, unordered multinomial Logistic regression analysis was used to examine the influencing factors of sleep subtypes in older adults with SCD. Results A total of 287 older adults with SCD were enrolled, and the results of the latent profile analysis showed that sleep in older adults with SCD can be classified into 3 potential subtypes: relatively good sleep subtype (n=200), sleep deprivation subtype (n=63), and difficulty falling asleep-medicated hypnosis subtype (n=24), accounting for 69.7%, 21.9%, and 8.4% of all respondents, respectively. There were significant differences in gender, smart phone use, PHQ-9 scores and FRAIL scores among different sleep subtypes (P<0.05). Using the relatively good sleep type as a reference, the unordered multinomial Logistic regression analysis showed that gender 〔sleep deprivation subtype: female, OR=2.479, 95%CI (1.279, 4.808) 〕, smart phone use 〔sleep deprivation subtype: yes, OR=0.269, 95%CI (0.090, 0.808) 〕, PHQ-9 score 〔sleep deprivation subtype: OR=1.755, 95%CI (1.416, 2.175); difficulty falling asleep-medicated hypnosis subtype: OR=1.992, 95%CI (1.540, 2.576) 〕were influencing factors of sleep subtyping (P<0.05) . Conclusion Sleep in older adults with SCD showed significant population heterogeneity, and more attention should be paid to the sleep status of older adults with SCD who are female, use smart phones, and have depressive tendencies. Early and precise interventions for different sleep subtypes need to be performed early to improve sleep quality and prevent or delay the development of cognitive impairment.

Keywords

• cognition disorders|subjective cognitive decline|sleep|sleep quality|latent profile analysis|root cause analysis|logistic regression