The age-specific comorbidity burden of mild cognitive impairment: a US claims database study

Area of Science:

• Neurology
• Geriatrics
• Data Science in Healthcare

Background:

• Early identification of mild cognitive impairment (MCI) patients at high risk for Alzheimer's disease and related dementia disorders (ADRD) is crucial for personalized prevention strategies.
• Investigating the link between MCI and ADRD comorbidities can enhance predictive capabilities.
• Machine learning algorithms can analyze complex comorbidity data for risk assessment.

Purpose of the Study:

• To investigate the association between MCI and comorbidities linked to ADRD.
• To evaluate the predictive power of these comorbidities for determining MCI risk.
• To develop a machine learning model for MCI risk prediction based on comorbidities.

Main Methods:

• A retrospective matched case-control study design was employed.
• Data from 5185 MCI and 15,555 non-MCI individuals aged 50 years and above were analyzed from MarketScan databases.
• Predictive models incorporated 25 ADRD comorbidities, age, and sex.

Main Results:

• Significant associations were found between 25 ADRD comorbidities and MCI, though these weakened with increasing age.
• Odds ratios for depression and stroke/transient ischemic attack in MCI vs. non-MCI individuals decreased with age.
• The predictive accuracy, measured by ROC-AUC, decreased in older age groups (0.75, 0.70, 0.66).

Conclusions:

• The comorbidity burden in individuals with MCI is significantly influenced by age.
• A predictive model utilizing comorbidities alone demonstrated reasonable accuracy in diagnosing MCI.
• Certain comorbidities serve as age-specific predictors for MCI.