In this study, we developed machine learning models for predicting mortality through training, testing, and validation using 163,782 visit records of 45,275 unique NACC individuals in the United States from 2005 to 2021. We have demonstrated that machine learning models, which have thus far primarily been explored as screening or diagnosis tools in the context of dementia, have substantial utility in the prediction of mortality among dementia patients. First, we conducted multiple survival analyses, which confirmed that increasing global CDR scores coincided with decreased survival and showed that there was considerable variability in survival across dementia subtypes. Subsequently, we developed two-feature models (using only age and standard global CDR) and multi-factorial models (using nine features determined through feature selection) to predict dementia patient mortality at four distinct survival-time thresholds, all of which achieved high predictive performance. We additionally built machine learning models for eight different dementia subtypes and revealed key feature differences among them, though age and cognitive features derived from neuropsychological tests remained important predictors of mortality across all dementia types. These mortality predictors reveal similarities and differences in the etiology and clinical representation among individuals affected by different types of dementia.
The results of our global CDR survival analysis were consistent with those of past survival analyses in dementia patients31,32, confirming that higher CDR scores correlate with reduced survival probability. With respect to dementia type, there have been very few studies investigating the association between dementia type and survival probability. In the studies that we identified, key distinctions were noted in the cohort composition and mortality risk of Lewy body dementia vs. Alzheimer’s disease27, vascular dementia vs. Alzheimer’s disease in the context of depression26, and among eight different dementia subtypes33. Our dementia-type survival analysis confirmed this heterogeneity, as survival probability differed drastically across groups of patients with different primary etiologic diagnoses. However, whereas prior studies identified comorbidities such as cardiovascular disease34 to be associated with reduced survival probability, we found that regardless of heart conditions, the survival curves separated decisively across patients with varied global CDR scores within the NACC cohort.
Subsequently, we built machine learning models tasked with predicting dementia patient mortality at one-, three-, five-, and ten-year survival thresholds. Our two-feature models, which utilized age and global CDR scores, achieved an AUC-ROC of over 0.76 at all four survival thresholds in the test set. Thus, age and global CDR provided a solid basis for predicting dementia patient mortality and, in the absence of additional clinical features, may alone be used to guide clinical judgment. Our multi-factorial models, for which we utilized SHAP to select a subset of nine features, achieved an AUC-ROC of over 0.82 at all four survival thresholds in the test set and comparable performance in the validation set. The crucial features used by the multi-factorial models confirm the known clinical indicators of dementia from a machine learning standpoint. The multi-factorial models revealed that a higher risk of mortality was predicted by older age35–38, male sex31,36–38, higher levels of dependency and personal care required38, more years of education39, more years of smoking40, and poorer performance on neuropsychological exams like the Trail Making Test41,42.
To our knowledge, our study is one of just a few studies to apply a machine learning-based approach to predicting mortality in dementia patients23–25 (as opposed to statistical approaches), and the first study to do so within population subsets stratified by dementia type. In predicting dementia patient mortality at the five-year survival threshold, our dementia type-specific models all achieved an AUC-ROC of over 0.79 in the test set and similar performance in the validation set. Hierarchical clustering of survival predictors grouped the following dementia types together: (1) vascular dementia (VaD) with depression, (2) Lewy body dementia (LBD) with frontotemporal lobar dementia (FTLD), (3) Alzheimer’s disease (AD) with other dementia, and (4) no dementia with unknown. Since many dementia types present similar symptoms and disease progressions8, differentiating and targeting dementia type-specific symptoms and mortality predictors can be beneficial for patient populations43. Across all four clusters (even in the no dementia and unknown cluster), many features from the multi-factorial models remained key predictors of mortality, such as age, level of independence, smoking, and performance on neuropsychological exams like the Trail Making Test.
First, within the VaD and depression cluster, body measurements and vital signs (e.g., height, weight, BMI, heart rate, and diastolic blood pressure) contributed to the mortality prediction more than for any other dementia type. For VaD, congestive heart failure was the second most important feature after age, consistent with VaD’s common risk factors8. Moreover, the grouping of VaD with depression confirms previous literature that has highlighted the synergistic effects of VaD and depression on patient mortality26, as VaD patients tend to exhibit a higher baseline risk for psychiatric symptoms like depression43,44. Second, within the FTLD and LBD cluster, features corresponding to MMSE score, standard CDR sum of boxes, and involvement in community affairs contributed more heavily to the mortality prediction. For FTLD in particular, features measuring difficulty in performing social and functional activities were the pivotal predictors of mortality, consistent with the pathological effects of FTLD8. Our findings regarding FTLD and LBD align with prior studies that have similarly grouped the two subtypes together and determined that executive dysfunction and activity disturbances are the key indicators of cognitive impairment for both43,45. Third, within the AD and other dementia clusters, general cognitive features, namely those from the multi-factorial models, remained the most important predictors of mortality. Standard CDR sum of boxes was also an important predictor of mortality in AD patients, as were body measurements and vital signs for other dementia patients. The grouping of AD with other dementia may be attributed to the difficulty in differentiating AD from certain other types of dementia46, and given that AD was by far the most prevalent dementia type in the NACC cohort, it is likely that the other dementia patients were generally similar to AD patients. Finally, within the no dementia and unknown cluster, general cognitive features such as performance on the Trail Making Test, surprisingly, remained important predictors of mortality. However, general comorbidities and mortality risk factors, such as smoking, hypertension, and lack of energy, demonstrated high relative importance as well, more so than for any of the dementias. Notably, as in the survival analysis, cardiovascular diseases did not appear in the top features in either the multi-factorial models or the dementia type-specific models, with the exception of congestive heart failure for VaD. The absence of these comorbidities from the top features in our machine learning models may suggest that cognitive decline is a stronger predictor of mortality in dementia patients than stroke or other comorbid cardiovascular conditions, though further studies could better interrogate this hypothesis.
Our study had several key strengths. First, the NACC database is the largest resource of its kind in the United States, covering a large, diverse patient population that was current through September 2021. Moreover, we highlight a conscious design choice in stratifying our data into train, test, and validation sets. By introducing a prospective validation set based on date, we were able to ascertain the ability of our models to predict mortality within a prospective cohort based on past data. In our pan-dementia analysis, the use of two-feature and nine-feature (multi-factorial) models provided a parsimonious, clinically feasible framework for predicting dementia patient mortality, while in our sub-dementia analysis, the comparison of important predictors of mortality across various dementia types may help to guide precision management and treatment of dementia.
However, our study also had limitations. Due to the high prevalence of missing values, largely attributed to the difficulty in acquiring certain data (e.g., neuropathological data) and differences in clinical procedures across ADCs, many features were preliminarily eliminated. Moreover, many features within the NACC data measure similar phenomena, certain variables have changed over time as updates were made to the UDS form, and many variables were derived from clinician diagnosis, precluding the use of a more granular feature selection method. By first eliminating variables with over 40% missing values and subsequently using MICE to impute the remaining features, we aimed to reduce some bias in the feature selection process47, though we acknowledge the limitation of neglecting features that may only be ascertained by a selection of ADCs. We highlight that the best performance can be achieved if each ADC or clinic derives its own predictive model based on its respective available features. Moreover, our data-splitting method excluded patients who are lost to follow-up, which biases the deceased group towards a shorter survival time. This will likely make the predictors over different survival thresholds more similar to each other and overestimate the AUC values for the longer survival thresholds.
Overall, this study revealed that machine learning models have utility in predicting dementia patient mortality at various survival-time thresholds. Parsimonious models can be developed when limited clinical features are available, and dementia type-specific models can be used for distinguishing heterogeneous patient populations. If cross-validated and carefully implemented at the primary care level, such predictive models can improve personalized care of dementia.