Multistate models and lifetime risk estimation: Application to Alzheimer's disease

Predictors of transitions between normal cognition, cognitive impairment, and dementia in a longitudinal cohort

BackgroundUnderstanding how sociodemographic characteristics and medical history are associated with progression (or regression) of Alzheimer's disease and related dementias could inform intervention strategies, personalized prognoses, and projections of population-level burden.ObjectiveWe estimated transition rates for progression and reversion between normal cognition, cognitive impairment, dementia, and death in a longitudinal cohort, as well as associations with sociodemographic characteristics and medical history.MethodsWe applied a multistate transition model to a cohort of 960 participants (with 2-16 (median 3) annual visits; 2006-24). Covariate hazard ratios (HRs) were estimated in models adjusted for age group.ResultsSeveral covariates were associated with faster progression from normal cognition to cognitive impairment but slower progression from cognitive impairment to dementia. For example, non-Hispanic Black participants transitioned from normal to cognitive impairment at higher rates (HR: 2.29, 95% CI: 1.63, 3.21) and to dementia at lower rates (HR: 0.12, 95% CI: 0.06, 0.23) than non-Hispanic White participants. Additionally, amnestic versus non-amnestic impairment emerged as a strong predictor of transitions from cognitive impairment by reducing reversion to normal cognition (HR: 0.51, 95% CI: 0.35, 0.74) and accelerating progression to dementia (HR: 2.51, 95% CI: 1.49, 4.22). History of traumatic brain injury was associated with reversion from cognitive impairment to normal cognition (HR: 2.43, 95% CI: 1.13, 5.23).ConclusionsA better understanding and measurement of cognitive impairment is needed to explain and predict both reversion to normal cognition and why factors associated with faster onset of impairment may be associated with delayed onset of dementia.

MSGene: a multistate model using genetic risk and the electronic health record applied to lifetime risk of coronary artery disease

Coronary artery disease (CAD) is the leading cause of death among adults worldwide. Accurate risk stratification can support optimal lifetime prevention. Current methods lack the ability to incorporate new information throughout the life course or to combine innate genetic risk factors with acquired lifetime risk. We designed a general multistate model (MSGene) to estimate age-specific transitions across 10 cardiometabolic states, dependent on clinical covariates and a CAD polygenic risk score. This model is designed to handle longitudinal data over the lifetime to address this unmet need and support clinical decision-making. We analyze longitudinal data from 480,638 UK Biobank participants and compared predicted lifetime risk with the 30-year Framingham risk score. MSGene improves discrimination (C-index 0.71 vs 0.66), age of high-risk detection (C-index 0.73 vs 0.52), and overall prediction (RMSE 1.1% vs 10.9%), in held-out data. We also use MSGene to refine estimates of lifetime absolute risk reduction from statin initiation. Our findings underscore our multistate model's potential public health value for accurate lifetime CAD risk estimation using clinical factors and increasingly available genetics toward earlier more effective prevention.

Estimating effects of time-varying exposures on mortality risk

Administrative databases have become an increasingly popular data source for population-based health research. We explore how mortality risk is associated with some health service utilization process via linked administrative data. A generalized Cox regression model is proposed using a time-dependent stratification variable to summarize lifetime service utilization. Recognizing the service utilization over time as an internal covariate in the survival analysis, conventional likelihood methods are inapplicable. We present an estimating function based procedure for estimating model parameters, and provide a testing procedure for updating the stratification levels. The proposed approach is examined both asymptotically and numerically via simulation. We motivate and illustrate the proposed approach using an on-going program pertaining to opioid agonist treatment (OAT) management for individuals identified with opioid use disorders. Our analysis of the OAT data indicates that the OAT effect on mortality risk decreases in successive OAT attempts, in which two risk classes based on an individual's treatment episode number are established: one with 1-3 OAT episodes, and the other with 4+ OAT episodes.

MSGene: Derivation and validation of a multistate model for lifetime risk of coronary artery disease using genetic risk and the electronic health record

Currently, coronary artery disease (CAD) is the leading cause of death among adults worldwide. Accurate risk stratification can support optimal lifetime prevention. We designed a novel and general multistate model (MSGene) to estimate age-specific transitions across 10 cardiometabolic states, dependent on clinical covariates and a CAD polygenic risk score. MSGene supports decision making about CAD prevention related to any of these states. We analyzed longitudinal data from 480,638 UK Biobank participants and compared predicted lifetime risk with the 30-year Framingham risk score. MSGene improved discrimination (C-index 0.71 vs 0.66), age of high-risk detection (C-index 0.73 vs 0.52), and overall prediction (RMSE 1.1% vs 10.9%), with external validation. We also used MSGene to refine estimates of lifetime absolute risk reduction from statin initiation. Our findings underscore the potential public health value of our novel multistate model for accurate lifetime CAD risk estimation using clinical factors and increasingly available genetics.

Progression from different blood glucose states to cardiovascular diseases: a prospective study based on multi-state model

AIMS

To quantify the trajectories from normoglycaemia to pre-diabetes, subsequently to type 2 diabetes mellitus (T2DM), cardiovascular diseases (CVD), and cardiovascular death, and the effects of risk factors on the rates of transition.

METHODS AND RESULTS

We used data from the Jinchang Cohort of 42 585 adults aged 20-88 free of coronary heart disease (CHD) and stroke at baseline. A multistate model was applied for analysing the progression of CVD and its relation to various risk factors. During a median follow-up of 7 years, 7498 participants developed pre-diabetes, 2307 developed T2DM, 2499 developed CVD, and 324 died from CVD. Among 15 postulated transitions, transition from comorbid CHD and stroke to cardiovascular death had the highest rate (157.21/1000 person-years), followed by transition from stroke alone to cardiovascular death (69.31/1000 person-years) and transition from pre-diabetes to normoglycaemia (46.51/1000 person-years). Pre-diabetes had a sojourn time of 6.77 years, and controlling weight, blood lipids, blood pressure, and uric acid within normal limits may promote reversion to normoglycaemia. Among transitions to CHD alone and stroke alone, transition from T2DM had the highest rate (12.21/1000 and 12.16/1000 person-years), followed by transition from pre-diabetes (6.81/1000 and 4.93/1000 person-years) and normoglycaemia (3.28/1000 and 2.39/1000 person-years). Age and hypertension were associated with an accelerated rate for most transitions. Overweight/obesity, smoking, dyslipidaemia, and hyperuricaemia played crucial but different roles in transitions.

CONCLUSION

Pre-diabetes was the optimal intervention stage in the disease trajectory. The derived transition rates, sojourn time, and influence factors could provide scientific support for the primary prevention of both T2DM and CVD.

The lifetime risk of developing type II diabetes in an urban community in Mumbai: findings from a ten-year retrospective cohort study

BACKGROUND

Incidence and prevalence do not capture the risk of developing diabetes during a defined period and only limited evidence exists on the lifetime risk of diabetes based on longer and continuous follow-up studies in India. Lacunae in evidence on lifetime risk can be attributed primarily to the absence of comprehensive and reliable information on diabetes incidence, mortality rates and lack of longitudinal studies in India. In light of the scarcity of evidence in India, the objective of this study was to estimate the incidence of diabetes and its lifetime risk in an urban community of Mumbai.

METHODS

The research study utilized data which is extracted from the electronic medical records of beneficiaries covered under the Contributory Health Service Scheme in Mumbai. The dataset included information on 1652 beneficiaries aged 40 years and above who were non-diabetic in 2011-2012, capturing their visit dates to medical center and corresponding laboratory test results over a span ten years from January, 2012- December, 2021. Survival analysis techniques are applied to estimate the incidence of diabetes. Subsequently, the remaining life years from the life table were utilized to estimate the lifetime risk of diabetes for each gender, stratified by age group.

RESULTS

A total of 546 beneficiaries developed diabetes in ten years, yielding an unadjusted incidence rate of 5.3 cases per 1000 person-years (95% CI: 4.9- 5.8 cases/ 1000 person years). The age-adjusted lifetime risk of developing type II diabetes in this urban community is estimated to be 40.3%. Notably, males aged 40 years and above had 41.5% chances of developing diabetes in their lifetime as compared to females with a risk of 39.4%. Moreover, the remaining lifetime risk of diabetes decreased with advancing age, ranging from 26.4% among 40-44 years old to 4.2% among those age 70 years and above.

CONCLUSION

The findings stress the significance of recognizing age specific lifetime risk and implementing early interventions to prevent or delay diabetes onset and to focus on diabetes management programs in India.

Forecasting prevalence and mortality of Alzheimer's disease using the partitioning models

OBJECTIVES

Health forecasting is an important aspect of ensuring that the health system can effectively respond to the changing epidemiological environment. Common models for forecasting Alzheimer's disease and related dementias (AD/ADRD) are based on simplifying methodological assumptions, applied to limited population subgroups, or do not allow analysis of medical interventions. This study uses 5 %-Medicare data (1991-2017) to identify, partition, and forecast age-adjusted prevalence and incidence-based mortality of AD as well as their causal components.

METHODS

The core underlying methodology is the partitioning analysis that calculates the relative impact each component has on the overall trend as well as intertemporal changes in the strength and direction of these impacts. B-spline functions estimated for all parameters of partitioning models represent the basis for projections of these parameters in future.

RESULTS

Prevalence of AD is predicted to be stable between 2017 and 2028 primarily due to a decline in the prevalence of pre-AD-diagnosis stroke. Mortality, on the other hand, is predicted to increase. In all cases the resulting patterns come from a trade-off of two disadvantageous processes: increased incidence and disimproved survival. Analysis of health interventions demonstrates that the projected burden of AD differs significantly and leads to alternative policy implications.

DISCUSSION

We developed a forecasting model of AD/ADRD risks that involves rigorous mathematical models and incorporation of the dynamics of important determinative risk factors for AD/ADRD risk. The applications of such models for analyses of interventions would allow for predicting future burden of AD/ADRD conditional on a specific treatment regime.

The Construction of a Multidomain Risk Model of Alzheimer's Disease and Related Dementias

BACKGROUND

Alzheimer's disease (AD) and related dementia (ADRD) risk is affected by multiple dependent risk factors; however, there is no consensus about their relative impact in the development of these disorders.

OBJECTIVE

To rank the effects of potentially dependent risk factors and identify an optimal parsimonious set of measures for predicting AD/ADRD risk from a larger pool of potentially correlated predictors.

METHODS

We used diagnosis record, survey, and genetic data from the Health and Retirement Study to assess the relative predictive strength of AD/ADRD risk factors spanning several domains: comorbidities, demographics/socioeconomics, health-related behavior, genetics, and environmental exposure. A modified stepwise-AIC-best-subset blanket algorithm was then used to select an optimal set of predictors.

RESULTS

The final predictive model was reduced to 10 features for AD and 19 for ADRD; concordance statistics were about 0.85 for one-year and 0.70 for ten-year follow-up. Depression, arterial hypertension, traumatic brain injury, cerebrovascular diseases, and the APOE4 proxy SNP rs769449 had the strongest individual associations with AD/ADRD risk. AD/ADRD risk-related co-morbidities provide predictive power on par with key genetic vulnerabilities.

CONCLUSION

Results confirm the consensus that circulatory diseases are the main comorbidities associated with AD/ADRD risk and show that clinical diagnosis records outperform comparable self-reported measures in predicting AD/ADRD risk. Model construction algorithms combined with modern data allows researchers to conserve power (especially in the study of disparities where disadvantaged groups are often grossly underrepresented) while accounting for a high proportion of AD/ADRD-risk-related population heterogeneity stemming from multiple domains.

Expanding the scope of health disparities research in Alzheimer's disease and related dementias: Recommendations from the "Leveraging Existing Data and Analytic Methods for Health Disparities Research Related to Aging and Alzheimer's Disease and Related Dementias" Workshop Series

Topics discussed at the "Leveraging Existing Data and Analytic Methods for Health Disparities Research Related to Aging and Alzheimer's Disease and Related Dementias" workshop, held by Duke University and the Alzheimer's Association with support from the National Institute on Aging, are summarized.  Ways in which existing data resources paired with innovative applications of both novel and well-known methodologies can be used to identify the effects of multi-level societal, community, and individual determinants of race/ethnicity, sex, and geography-related health disparities in Alzheimer's disease and related dementia are proposed.  Current literature on the population analyses of these health disparities is summarized with a focus on identifying existing gaps in knowledge, and ways to mitigate these gaps using data/method combinations are discussed at the workshop.  Substantive and methodological directions of future research capable of advancing health disparities research related to aging are formulated.

Multistate Markov model application for blood pressure transition among the Chinese elderly population: a quantitative longitudinal study

OBJECTIVE

To explore the transitions of different blood pressure states based on a multistate Markov model among the Chinese elderly population.

SETTING

A community health centre in Xiamen, China.

PARTICIPANTS

1833 elderly Chinese people.

METHODS

A multistate Markov model was built based on 5001 blood pressure measurements from 2015 to 2020. Research was conducted to explore the process of hypertension progression, providing information on the transition probability, HR and the mean sojourn time in three blood pressure states, namely normal state, elevated state and hypertensive state.

RESULTS

Probabilities of moving from the normal state to the hypertensive state in the first year were 16.97% (female) and 21.73% (male); they increased dramatically to 47.31% (female) and 51.70% (male) within a 3-year follow-up period. The sojourn time in the normal state was 1.5±0.08 years. Elderly women in the normal state had a 16.97%, 33.30% and 47.31% chance of progressing to hypertension within 1, 2 and 3 years, respectively. The corresponding probabilities for elderly men were 21.73%, 38.56% and 51.70%, respectively. For elderly women starting in the elevated state, the probabilities of developing hypertension were 25.07%, 43.03% and 56.32% in the next 1, 2 and 3 years, respectively; while the corresponding changes for elderly men were 20.96%, 37.65% and 50.86%. Increasing age, body mass index (BMI) and glucose were associated with the probability of developing hypertension from the normal state or elevated state.

CONCLUSIONS

Preventive actions against progression to hypertension should be conducted at an early stage. More awareness should be paid to elderly women with elevated state and elderly men with normal state. Increasing age, BMI and glucose were critical risk factors for developing hypertension. The derived transition probabilities and sojourn time can serve as a significant reference for making targeted interventions for hypertension progression among the Chinese elderly population.

A comparison of statistical methods to predict the residual lifetime risk

Lifetime risk measures the cumulative risk for developing a disease over one's lifespan. Modeling the lifetime risk must account for left truncation, the competing risk of death, and inference at a fixed age. In addition, statistical methods to predict the lifetime risk should account for covariate-outcome associations that change with age. In this paper, we review and compare statistical methods to predict the lifetime risk. We first consider a generalized linear model for the lifetime risk using pseudo-observations of the Aalen-Johansen estimator at a fixed age, allowing for left truncation. We also consider modeling the subdistribution hazard with Fine-Gray and Royston-Parmar flexible parametric models in left truncated data with time-covariate interactions, and using these models to predict lifetime risk. In simulation studies, we found the pseudo-observation approach had the least bias, particularly in settings with crossing or converging cumulative incidence curves. We illustrate our method by modeling the lifetime risk of atrial fibrillation in the Framingham Heart Study. We provide technical guidance to replicate all analyses in R.

Senolytic Therapy to Modulate the Progression of Alzheimer's Disease (SToMP-AD): A Pilot Clinical Trial

Preclinical studies indicate an age-associated accumulation of senescent cells across multiple organ systems. Emerging evidence suggests that tau protein accumulation, which closely correlates with cognitive decline in Alzheimer's disease and other tauopathies, drives cellular senescence in the brain. Pharmacologically clearing senescent cells in mouse models of tauopathy reduced brain pathogenesis. Compared to vehicle treated mice, intermittent senolytic administration reduced tau accumulation and neuroinflammation, preserved neuronal and synaptic density, restored aberrant cerebral blood flow, and reduced ventricular enlargement. Intermittent dosing of the senolytics, dasatinib plus quercetin, has shown an acceptable safety profile in clinical studies for other senescence-associated conditions. With these data, we proposed and herein describe the objectives and methods for a clinical vanguard study. This initial open-label clinical trial pilots an intermittent senolytic combination therapy of dasatinib plus quercetin in five older adults with early-stage Alzheimer's disease. The primary objective is to evaluate the central nervous system penetration of dasatinib and quercetin through analysis of cerebrospinal fluid collected at baseline and after 12 weeks of treatment. Further, through a series of secondary outcome measures to assess target engagement of the senolytic compounds and Alzheimer's disease-relevant cognitive, functional, and physical outcomes, we will collect preliminary data on safety, feasibility, and efficacy. The results of this study will be used to inform the development of a randomized, double-blind, placebo-controlled multicenter phase II trial to further explore of the safety, feasibility, and efficacy of senolytics for modulating the progression of Alzheimer's disease. Clinicaltrials.gov registration number and date: NCT04063124 (08/21/2019).

Spillover: The Approval of New Medications for Alzheimer's Disease Dementia Will Impact Biomarker Disclosure Among Asymptomatic Research Participants

In this article we address how the recent, and anticipated upcoming, FDA approvals of novel anti-amyloid medications to treat individuals with mild Alzheimer's disease (AD) dementia could impact disclosure of biomarker results among asymptomatic research participants. Currently, research is typically the context where an asymptomatic individual may have the option to learn their amyloid biomarker status. Asymptomatic research participants who learn their amyloid status may have questions regarding the meaning of this result and the implications for accessing a potential intervention. After outlining our rationale, we provide examples of how current educational materials used in research convey messages regarding amyloid positivity and the availability of treatments, or lack thereof. We suggest language to improve messaging, as well as strengths of current materials, in addressing these issues for research participants. Although novel medications are currently only approved for use among symptomatic individuals, their availability may have implications for disclosure among asymptomatic research participants with evidence of amyloid deposition, who may be especially interested in information on these interventions for potential prevention, or future treatment, of mild cognitive impairment or dementia due to AD.

Alzheimer's disease neuropathology is exacerbated following traumatic brain injury. Neuroprotection by co-administration of nanowired mesenchymal stem cells and cerebrolysin with monoclonal antibodies to amyloid beta peptide

Military personnel are prone to traumatic brain injury (TBI) that is one of the risk factors in developing Alzheimer's disease (AD) at a later stage. TBI induces breakdown of the blood-brain barrier (BBB) to serum proteins into the brain and leads to extravasation of plasma amyloid beta peptide (ΑβP) into the brain fluid compartments causing AD brain pathology. Thus, there is a need to expand our knowledge on the role of TBI in AD. In addition, exploration of the novel roles of nanomedicine in AD and TBI for neuroprotection is the need of the hour. Since stem cells and neurotrophic factors play important roles in TBI and in AD, it is likely that nanodelivery of these agents exert superior neuroprotection in TBI induced exacerbation of AD brain pathology. In this review, these aspects are examined in details based on our own investigations in the light of current scientific literature in the field. Our observations show that TBI exacerbates AD brain pathology and TiO₂ nanowired delivery of mesenchymal stem cells together with cerebrolysin-a balanced composition of several neurotrophic factors and active peptide fragments, and monoclonal antibodies to amyloid beta protein thwarted the development of neuropathology following TBI in AD, not reported earlier.

Blood Pressure States Transition Inference Based on Multi-State Markov Model

The investigation of risk factors associated with hypertension patients has been extensively studied in the past decades. However, the pattern of natural progressive trajectories to hypertension from nonhypertensive states was rarely explored. In this study, we are interested in discovering the underlying transition patterns between different blood pressure states, namely normal state, elevated state, and hypertensive state among the working population in the United States. A multi-state Markov model was built based on 88,966 clinical records from 34,719 participants we collected during the worksite preventive screening from 2012 to 2018. We first investigated the various risk factors, and we found that body mass index (BMI) is the most critical factor for developing new-onset hypertension. The transition probabilities, survival probabilities, and sojourn time of each state were derived given different levels of BMI, age groups, and gender categories. We found the underweight participants are more likely to remain in the current nonhypertensive states within 3 years, while extremely obese participants have a higher probability of developing hypertension. We discovered the distinct transition patterns among male and female participants. On average, the sojourn time in the normal state for normal-weight participants is 4.33 years for females and 2.18 years for their male counterparts. For the extremely obese participants, the average sojourn time in the normal state is 1.38 years for females and 0.71 years for males. In the end, a web-based graphical user interface (GUI) application was developed for clinicians to visualize the impact of behavioral interventions on delaying the progression of hypertension. Our analysis can provide a unique insight into hypertension research and proactive interventions.