PhD student, Radboud University Medical Center
Understanding and Predicting Personalized Growth Rate curves to Aid Individualized care in Dementia
This project focuses on the progression of dementia. The progression of dementia at old age is very heterogeneous and reliable predictions of an individual’s outlook are therefore notoriously difficult to make. This leads to great uncertainty for patients as well as their relatives and physicians, and is a huge limitation for personalized care planning and shared decision making. My study aims to predict future progression, based on personal characteristics, ranging from physical health and frailty to supportive social networks. Dementia affects not only the cognitive abilities of patients, but also their ability to perform daily life activities, like eating and dressing. Dependency in activities of daily life is one of the greatest fears of older adults, diminishing their quality of life and increasing demands on family members, caregivers and care services. Therefore, this is one of the main outcomes of my study, along with cognitive performance. Understanding why certain dementia patients decline faster than others can aid advanced care planning by providing a personalized prognosis. Moreover, it can provide insights into new ways of preventing decline in dementia.
Abstract: Knowledge on survival in dementia is crucial for patients and public health planning. Most studies comparing mortality risk included few different dementia diagnoses.To compare mortality risk in the most frequent dementia disorders in a large cohort of patients with an incident diagnosis, adjusting for potential confounding factors.15,209 patients with dementia from the national quality database, Swedish Dementia Registry (SveDem), diagnosed in memory clinics from 2008 to 2011, were included in this study. The impact of age, gender, dementia diagnosis, baseline Mini-Mental State Examination (MMSE), institutionalization, coresidency, and medication on survival after diagnosis were examined using adjusted hazard ratios (HR) with 95% confidence intervals (CI).During a mean follow-up of 2.5 years, 4,287 deaths occurred, with 114 (95% CI 111-117) deaths/1,000 person-years. Adjusted HR of death for men was 1.56 (95% CI 1.46-1.66) compared to women. Low MMSE, institutionalization, and higher number of medications were associated with higher HR of death. All dementia diagnoses demonstrated higher HR compared to Alzheimer's disease, with vascular dementia presenting the highest crude HR. After adjusting, frontotemporal dementia had the highest risk with a HR of 1.91 (95% CI 1.52-2.39), followed by Lewy body dementia (HR 1.64; 95% CI 1.39-1.95), vascular dementia (HR 1.55; 95% CI 1.42-1.69), Parkinson's disease dementia (HR 1.47; 95% CI 1.17-1.84), and mixed Alzheimer's disease and vascular dementia (HR 1.32; 95% CI 1.22-1.44).Worse cognition, male gender, higher number of medications, institutionalization, and age were associated with increased death risk after dementia diagnosis. Adjusted risk was lowest in Alzheimer's disease patients and highest in frontotemporal dementia subjects.
Pub.: 15 Mar '14, Pinned: 27 Aug '17
Abstract: Several longitudinal studies of Alzheimer's disease (AD) report heterogeneity in progression. We sought to identify groups (classes) of progression trajectories in the population-based Cache County Dementia Progression Study (N = 328) and to identify baseline predictors of membership for each group.We used parallel-process growth mixture models to identify latent classes of trajectories on the basis of Mini-Mental State Exam (MMSE) and Clinical Dementia Rating sum of boxes scores over time. We then used bias-corrected multinomial logistic regression to model baseline predictors of latent class membership. We constructed receiver operating characteristic curves to demonstrate relative predictive utility of successive sets of predictors.We fit four latent classes; class 1 was the largest (72%) and had the slowest progression. Classes 2 (8%), 3 (11%), and 4 (8%) had more rapid worsening. In univariate analyses, longer dementia duration, presence of psychosis, and worse baseline MMSE and Clinical Dementia Rating sum of boxes were associated with membership in class 2, relative to class 1. Lower education was associated with membership in class 3. In the multivariate model, only MMSE remained a statistically significant predictor of class membership. Receiver operating characteristic areas under the curve were 0.98, 0.88, and 0.67, for classes 2, 3, and 4 relative to class 1.Heterogeneity in AD course can be usefully characterized using growth mixture models. The majority belonged to a class characterized by slower decline than is typically reported in clinical samples. Class membership could be predicted using baseline covariates. Further study may advance our prediction of AD course at the population level and in turn shed light on the pathophysiology of progression.
Pub.: 05 Nov '14, Pinned: 27 Aug '17
Abstract: Co-occurrence with other chronic diseases may influence the progression of dementia, especially in case of multiple chronic diseases. We aimed to verify whether multimorbidity influenced cognitive and daily functioning during nine years after dementia diagnosis compared with the influence in persons without dementia.In the Kungsholmen Project, a population-based cohort study, we followed 310 persons with incident dementia longitudinally. We compared their trajectories with those of 679 persons without dementia. Progression was studied for cognition and activities of daily life (ADLs), measured by MMSE and Katz Index respectively. The effect of multimorbidity and its interaction with dementia status was studied using individual growth models.The mean (SD) follow-up time was 4.7 (2.3) years. As expected, dementia related to both the decline in cognitive and daily functioning. Irrespective of dementia status, persons with more diseases had significantly worse baseline daily functioning. In dementia patients having more diseases also related to a significantly faster decline in daily functioning. Due to the combination of lower functioning in ADLs at baseline and faster decline, dementia patients with multimorbidity were about one to two years ahead of the decline of dementia patients without any co-morbidity. In persons without dementia, no significant decline in ADLs over time was present, nor was multimorbidity related to the decline rate. Cognitive decline measured with MMSE remained unrelated to the number of diseases present at baseline.Multimorbidity was related to baseline daily function in both persons with and without dementia, and with accelerated decline in people with dementia but not in non-demented individuals. No relationship of multimorbidity with cognitive functioning was established. These findings imply a strong interconnection between physical and mental health, where the greatest disablement occurs when both somatic and mental disorders are present.
Pub.: 05 Jan '14, Pinned: 27 Aug '17
Abstract: Alzheimer's disease is a neurodegenerative syndrome characterized by multiple dimensions including cognitive decline, decreased daily functioning and psychiatric symptoms. This systematic review aims to investigate the relation between somatic comorbidity burden and progression in late-onset Alzheimer's disease (LOAD).We searched four databases for observational studies that examined cross-sectional or longitudinal associations of cognitive or functional or neuropsychiatric outcomes with comorbidity in individuals with LOAD. From the 7966 articles identified originally, 11 studies were included in this review. The Newcastle-Ottawa quality assessment was used. The large variation in progression measures, comorbidity indexes and study designs hampered the ability to perform a meta-analysis. This review was registered with PROSPERO under DIO: 10.15124/CRD42015027046.Nine studies indicated that comorbidity burden was associated with deterioration in at least one of the three dimensions of LOAD examined. Seven out of ten studies investigating cognition found comorbidities to be related to decreased cognitive performance. Five out of the seven studies investigating daily functioning showed an association between comorbidity burden and decreased daily functioning. Neuropsychiatric symptoms (NPS) increased with increasing comorbidity burden in two out of three studies investigating NPS. Associations were predominantly found in studies analyzing the association cross-sectionally, in a time-varying manner or across short follow-up (≤2 years). Rarely baseline comorbidity burden appeared to be associated with outcomes in studies analyzing progression over longer follow-up periods (>2 years).This review provides evidence of an association between somatic comorbidities and multifaceted LOAD progression. Given that time-varying comorbidity burden, but much less so baseline comorbidity burden, was associated with the three dimensions prospectively, this relationship cannot be reduced to a simple cause-effect relation and is more likely to be dynamic. Therefore, both future studies and clinical practice may benefit from regarding comorbidity as a modifiable factor with a possibly fluctuating influence on LOAD.
Pub.: 05 May '17, Pinned: 24 Aug '17
Abstract: Dementia is a neurodegenerative syndrome that interferes with multiple aspects of life, including cognition, daily functioning, and behavior. Despite the large heterogeneity in symptom development, these three domains are seldom studied simultaneously. This study investigates how trajectories of these domains are interrelated within individuals over time, and how they in turn are related to dementia severity and quality of life (QoL).We used data from a longitudinal clinical cohort study, including 331 dementia patients. Cognitive status was measured using the Mini-Mental State Examination, daily functioning was measured with the disability assessment for dementia and neuropsychiatric symptoms (NPS) were scored using the neuropsychiatric inventory. We investigated the relationships in the time course of the various dementia domains using random effects multilevel models and parallel-process growth models.Changes in cognition and daily functioning were highly correlated over time (r = 0.85, p < 0.01), as were changes in NPS and functioning (r = -0.60, p < 0.01), while changes in cognition and NPS were not (r = -0.20, p = 0.06). All three domains were strongly associated with dementia severity over time (p < 0.01). Decreased functioning and increased NPS were both associated with decreased QoL (β = 2.97, p < 0.01 and β = -2.41, p < 0.01, respectively), while cognition was not (β = 0.01, p = 0.93).This study demonstrates the heterogeneity of dementia progression between individuals and between different dementia domains within individuals. To improve our understanding of dementia progression, future research should embrace a broader perspective encompassing multiple outcome measures along with the patient's profile, including neurological factors as well as physical, social, and psychiatric health.
Pub.: 14 Mar '17, Pinned: 24 Aug '17