Dalhousie University


Research / Projects
Multi-State Transition Model

On average, cognition declines with age, but it declines more rapidly in some people than in others. In some individuals, cognition can even improve, although this is sometimes viewed as a result of diagnostic unreliability, and so has received less attention.

Widely used statistical models focus on dichotomous outcomes, (e.g. worse/not worse, alive/dead). However multi-state transition models allow simultaneous consideration of many outcomes. For example, this model can be used to estimate the impact of common covariates on individual probabilities of cognitive improvement, stability, decline and death.

The figure below illustrates the general schema of transitions between the different states (Si, i = 0,1,…,N) including death (D).

Recently at the Geriatric Medicine Research Unit, a new extension of multi-state transition models has been developed for comprehensive analyses of longitudinal data. It is based on a new appreciation that the brain in Alzheimer’s disease is not passive; it fights back by compensating. This compensation is dynamic – factors interact and change over time. These factors can be taken into account with dynamic system modeling.

In our previous research, this analytical technique was successfully applied to analysing general health status during aging, as well as its cognitive dimension. It allows analysis of changes in all directions, including improvement – which has had relatively little attention, but appears to be essential to understanding how AD occurs. Indeed, further development of the model and unknown risk factors associated with age-related improvement is the subject of ongoing inquiries by our group. 


Further Information:

If you have any questions, please contact Nader Fallah Tel: (902) 473-3350


Further Reading

Going from bad to worse: a stochastic model of transitions in deficit accumulation, in relation to mortality. Mitnitski A, Bao L, Rockwood K.  Mechanisms of Ageing and Development 2006; 127:490-493.

A cross-national study of transitions in deficit counts in two birth cohorts: implications for modeling ageing. Mitnitski A, Bao L, Skoog I, Rockwood K. Experimental Gerontology 2007; 42:241-246. 

Improvement and decline in health status from late middle age: Modeling age-related changes in deficit accumulation. Mitnitski A, Song X, Rockwood K. Experimental Gerontology 2007; 42:1109-1115.

Transitions in cognitive test scores over 5 and 10 years in elderly people: evidence for a model of age-related deficit accumulation. Mitnitski A, Rockwood K. BMC Geriatr 2008; 8:3.

Changes in cognition and mortality in relation to exercise in late life: a population based study. Middleton LE, Mitnitski A, Fallah N, Kirkland SA, Rockwood K. PloS ONE 2008; 3:e3,124.

Modeling the Impact of Sex on How Exercise Is Associated with Cognitive Changes and Death in Older Canadians. Fallah N, Mitnitski A, Middleton L, Rockwood K. Neuroepidemiology. 2009; 8:33(1):47-54.



Dalhousie University Faculty of Medicine