Identification of risk factors and development of a predictive nomogram for sarcopenia in Alzheimer's disease

Abstract

Introduction: Sarcopenia, with its complex diagnostic process, is a likely independent predictor of poor prognosis in patients with Alzheimer's disease (AD). However, research on the clinical characteristics and biomarkers of AD patients with sarcopenia (ADSA) is limited.

Methods: This study included 180 ADSA and 188 AD patients without sarcopenia (ADNSA), and evaluated demographics, cognitive function, motor capacity, emotional state, and daily living abilities.

Results: ADSA patients were older, with worse motor and cognitive functions, more severe depression, poorer social functioning, and lower daily living abilities compared to ADNSA patients. Multivariate regression identified age, low Frailty Rating Scale (FRS) scores, low serum albumin level, and low creatinine/cystatin C ratio (CCR) as risk factors for sarcopenia. A nomogram model based on these indicators demonstrated high discriminative power and clinical utility.

Discussion: Sarcopenia significantly affects AD patients' various functions. The nomogram model aids in the early detection of and personalized interventions for sarcopenia in AD.

Highlights: Sarcopenia is a risk factor for Alzheimer's disease (AD), and the coexistence of sarcopenia affects various functions and quality of life in patients with AD. Serum albumin and Frailty Rating Scale (FRS) scores are significantly associated with both sarcopenia and cognitive assessment indicators in AD patients with sarcopenia (ADSA). The combined sarcopenia nomogram model with indexes of age at diagnosis, creatinine/cystatin C ratio (CCR), FRS score, and albumin levels can aid in effectively identifying and personalizing interventions for sarcopenia in the AD population.

Keywords: Alzheimer's disease; biomarkers; nomogram; prediction model; risk factors; sarcopenia.

FIGURE 1

A nomogram predicting the risk of sarcopenia for patients with Alzheimer's disease. The value of each variable was given a score on the points scale axis. A total point could be easily calculated by adding each single score, and by projecting the total score to the lower total point scale, we were able to estimate the precise probability of sarcopenia. For example, a 75‐year‐old AD patient with an FRS score of 4, a serum albumin level of 40 g/L, and a CCR of 80, has points of 63.75, 51.25, 52.5, and 26.25, respectively. The total points is the sum of all the points, which is 193.75, and the corresponding risk of the event is ≈0.91, suggesting a 91% risk of sarcopenia in this patient. AD, Alzheimer's disease; CCR, creatinine/cystatin C ratio; FRS, Frailty Rating Scale.

FIGURE 2

The ROC cure of the nomogram model and each single index. The AUC values for the individual predictors were shown in the figure, and the nomogram model with combined index had the highest AUC of 0.822. AUC, area under the curve; ROC, receiver‐operating characteristic.

FIGURE 3

(A) The calibration curves for the nomogram. The x‐axis represents the nomogram‐predicted probability and y‐axis represents the actual probability of sarcopenia of AD patients. Perfect prediction would correspond to the black solid line. The green dotted line represents the entire cohort (n = 368), and the red solid line is bias‐corrected by bootstrapping (B = 1000 repetitions), indicating observed nomogram performance. (B) The DCA curves for the nomogram. The red solid line showed all patients being diagnosed with AD, and the green solid line showed none of patients being diagnosed with AD. AD, Alzheimer's disease; DCA, decision curve analysis.