Abstract
Background
Alzheimer’s disease (AD) is the leading cause of dementia with the incidence of this disease predicted to increase at least three fold by 2050. Curing this disease is a global priority. To achieve this an understanding of the entire disease course is necessary. However, data available on AD progression is quite sparse. Typically available data (AIBL and ADNI) only has around 5 years of follow-up and minimal information about disease stage. Previous research has suggested that progression from healthy to AD could span approximately 30 years1and postulates that disease markers will follow a sigmoidal shape2. However, little empirical evidence is available to support these findings.
Methods
To reconstruct and quantify the underlying longitudinal trajectories for disease markers were evaluated using individuals’ short-term follow-up data. Simulated sparse follow-up data for a number of individuals was created in order to investigate possible solutions, with the least deviation from the underlying simulated model considered most accurate. A four-step modeling approach was adopted that 1) determined individual slopes and anchor points, 2) fitted polynomials to the reciprocated data, 3) integrated and 4) inverted the fitted polynomial to obtain the longitudinal trajectories. Variations in the approach were conducted to determine the optimal method: mixed models were contrasted to linear regression in step 1) and different orders of polynomials were considered in step 2).
Results
Longitudinal trajectories of possible disease markers (Figure 1) were successfully reproduced from the sparse individual data. All methods accurately reproduced the underlying curve (0-3.6% difference between the simulated and predicted curves). The mixed model variations performed better than those using linear regression and there were minimal differences between the different order polynomials.
Conclusions
Simple modeling techniques are able to reconstruct generalized underlying longitudinal trajectories from individuals’ short-term follow-up data. The elucidation of these generalized underlying progression curves allow the timing associated with disease progression to be determined, providing imperative information for a wide range of applications including the design and timing of therapeutic interventions and planning for events such as residential care needs.
1 10.1016/S1474-4422(13)70044-9
2 10.1016/S1474-4422(09)70299-6
Alzheimer’s disease (AD) is the leading cause of dementia with the incidence of this disease predicted to increase at least three fold by 2050. Curing this disease is a global priority. To achieve this an understanding of the entire disease course is necessary. However, data available on AD progression is quite sparse. Typically available data (AIBL and ADNI) only has around 5 years of follow-up and minimal information about disease stage. Previous research has suggested that progression from healthy to AD could span approximately 30 years1and postulates that disease markers will follow a sigmoidal shape2. However, little empirical evidence is available to support these findings.
Methods
To reconstruct and quantify the underlying longitudinal trajectories for disease markers were evaluated using individuals’ short-term follow-up data. Simulated sparse follow-up data for a number of individuals was created in order to investigate possible solutions, with the least deviation from the underlying simulated model considered most accurate. A four-step modeling approach was adopted that 1) determined individual slopes and anchor points, 2) fitted polynomials to the reciprocated data, 3) integrated and 4) inverted the fitted polynomial to obtain the longitudinal trajectories. Variations in the approach were conducted to determine the optimal method: mixed models were contrasted to linear regression in step 1) and different orders of polynomials were considered in step 2).
Results
Longitudinal trajectories of possible disease markers (Figure 1) were successfully reproduced from the sparse individual data. All methods accurately reproduced the underlying curve (0-3.6% difference between the simulated and predicted curves). The mixed model variations performed better than those using linear regression and there were minimal differences between the different order polynomials.
Conclusions
Simple modeling techniques are able to reconstruct generalized underlying longitudinal trajectories from individuals’ short-term follow-up data. The elucidation of these generalized underlying progression curves allow the timing associated with disease progression to be determined, providing imperative information for a wide range of applications including the design and timing of therapeutic interventions and planning for events such as residential care needs.
1 10.1016/S1474-4422(13)70044-9
2 10.1016/S1474-4422(09)70299-6
| Original language | English |
|---|---|
| Pages (from-to) | P867-P868 |
| Number of pages | 2 |
| Journal | Alzheimer's and Dementia |
| Volume | 11 |
| Issue number | 7 (Supplement) |
| DOIs | |
| Publication status | Published - Jul 2015 |