Part II: A step-by-step guide to latent class analysis

Aflaki, Kayvan; Vigod, Simone N.; Ray, Joel G.

doi:10.1016/j.jclinepi.2023.05.025

Cited by 6 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Latent class analysis was performed with Mplus to identify the latent classes of IC impairment. We reported model fit indices for each model, such as the value of the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Sample-size Adjusted BIC (SABIC), Entropy, and Lo-Mendell-Rubin (LMR) [ 38 ]. Lower AIC, BIC and SABIC values and higher entropy indicate better model fit.…”

Section: Methodsmentioning

confidence: 99%

Factors associated with intrinsic capacity impairment in hospitalized older adults: a latent class analysis

Zhu,

Shen,

Shi

et al. 2024

BMC Geriatr

View full text Add to dashboard Cite

Background Intrinsic capacity (IC) is proposed by the World Health Organization (WHO) to promote healthy aging. Although some studies have examined the factors influencing IC, few studies have comprehensively confirmed lifestyle factors on IC, especially IC impairment patterns. The present study aimed to identify the patterns of IC impairment and explore the lifestyle and other factors associated with different patterns of IC impairment. Methods This cross-sectional study was conducted in a Chinese geriatric hospital. IC was evaluated in five domains according to the recommendations of WHO: cognition, locomotion, vitality, sensory and psychological domains. The sociodemographic and health-related characteristics of participants were assessed.The health promoting lifestyle was evaluated using the Health-Promoting Lifestyle Profile-II scale, including nutrition, health responsibility, interpersonal relationships, physical activity, spiritual growth and stress management. We applied latent class analysis to identify IC impairment patterns and compared basic activities of daily living, instrumental activities of daily living, frailty, quality of life and falls among different IC impairment patterns. Multinomial logistic regression analysis was conducted to identify factors influencing the IC impairment patterns. Results Among 237 participants included, the latent class analysis identified three patterns of IC impairment: 44.7% high IC (Class 1), 31.2% intermediate IC mainly locomotor impairment (Class 2) and 24.1% low IC mainly cognitive impairment (Class 3). Older adults in class 1 had the best function ability and quality of life, while class 3 had the highest levels of disability and frailty, the poorest quality of life and a higher prevalence of falls. Compared with class 1, older adults with advanced age (OR = 22.046, 95%CI:1.735-280.149), osteoporosis (OR = 3.377, 95%CI:1.161–9.825), and lower scores in physical activity (OR = 0.842, 95%CI:0.749–0.945), stress management (OR = 0.762, 95%CI:0.585–0.993) and social support (OR = 0.897, 95%CI:0.833–0.965) were more likely to belong to the class 2. Simultaneously, compared with class 1, older adults with advanced age (OR = 104.435, 95%CI:6.038-1806.410), stroke (OR = 3.877, 95%CI:1.172–12.823) and lower scores in physical activity (OR = 0.784, 95%CI:0.667–0.922) and social support (OR = 0.909, 95%CI:0.828–0.998) were more likely to be class 3. In addition, compared with class 2, older adults with a lower score in nutrition (OR = 0.764, 95%CI:0.615–0.950) were more likely to belong to the class 3. Conclusions This study provides evidence that there are heterogeneous IC impairment patterns in older adults and identifies various associated factors in each pattern, including age, stroke, osteoporosis, social support and lifestyle behaviors such as nutrition, physical activity and stress management. It informs stakeholders on which modifiable factors should be targeted through public health policy or early intervention to promote IC and healthy aging in older adults.

show abstract

Section: Methodsmentioning

confidence: 99%

Factors associated with intrinsic capacity impairment in hospitalized older adults: a latent class analysis

Zhu,

Shen,

Shi

et al. 2024

BMC Geriatr

View full text Add to dashboard Cite

show abstract

“…LCA is a statistical approach to uncovering unobserved subgroups (i.e., "latent classes") that exist within a population. For more information, the reader is encouraged to read the user-friendly introductions to LCA by Aflaki et al 34,35 . We had previously utilized LCA and WOMAC scores to identify 2 subgroups of patients with different outcomes within a cohort that underwent arthroplasty 36 .…”

Section: Alternatives To Mcidsmentioning

confidence: 99%