Sequential diagnostic patterns may offer more insight into Alzheimer’s disease risk than individual conditions alone.
In a retrospective study analyzing medical records from 24,473 patients, researchers identified 4 common diagnostic trajectories that frequently preceded Alzheimer disease (AD). These multistep sequences—centered on mental health disorders, encephalopathy, mild cognitive impairment (MCI), and vascular disease—were associated with higher risk for AD than isolated diagnoses.
Using more than a decade of data from the University of California Health Data Warehouse, the investigators examined the sequence and timing of diagnoses preceding AD. Of 5,762 patients with multistep trajectories, the team mapped 6,794 unique diagnostic sequences composed of 2 to 9 steps. Patients were aged 65–90 years, and the analysis employed a Fine–Gray subdistribution hazard model to account for death as a competing risk.
The researchers applied dynamic time warping and k-means clustering to group similar sequences, resulting in four distinct trajectory clusters:
-
In the mental health cluster, sequences commonly began with hypertension and depressive episodes.
-
The encephalopathy cluster featured unspecified brain disorders following conditions such as hypertension or kidney disease.
-
The MCI cluster often included transient cerebral ischemic attacks or hormonal disorders preceding cognitive impairment.
-
The vascular cluster, centered on cerebrovascular disease, frequently included joint disorders, anemia, or soft tissue disorders.
Despite differences in path composition, a key finding was that sequence order mattered. Patients who followed specific trajectories—such as “hypertension → depression → AD”—had higher risk than those with the same diagnoses in a different order. Causal inference modeling found that 26.2% of diagnostic transitions showed consistent directional patterns, with the highest proportion in the encephalopathy cluster.
Demographic characteristics differed by cluster. Patients in the mental health group were more likely to be women and Hispanic, while the vascular group had longer medical histories and higher comorbidity burden. The encephalopathy cluster showed the shortest interval between first diagnosis and AD, and the shortest survival following AD diagnosis.
To test reproducibility, the team validated findings in the All of Us Research Program, a national cohort of more than 500,000 individuals. Among patients with AD in this cohort, 89.9% were assignable to one of the four original clusters. In this validation set, 7 of 9 tested multistep trajectories remained significantly associated with AD risk, even after adjusting for demographic and clinical covariates.
According to the authors, these findings suggest that diagnostic history—and the order of diagnoses—may serve as a clinically meaningful predictor of AD risk. Identifying such trajectories could support earlier detection and personalized intervention strategies.
The authors reported no conflicts of interest.
Source: Science Direct