Background Dementia is a highly disabling condition causing considerable economic and societal burden worldwide. Predictions of future dementia prevalence need to take into account that, on the one hand, cardiovascular disease risk reduction may result in reduced age-specific dementia incidence but, on the other hand, increased life expectancy may lead to a larger pool of individuals susceptible to dementia. To accurately forecast prevalence of dementia, we modelled simultaneously the two conditions.

Methods We developed IMPACT-BAM to model transitions to 2030 in the England and Wales population, aged 35+, through health states of cardiovascular disease, cognitive and functional impairment, and dementia to cardiovascular and non-cardiovascular death. At each iteration of this probabilistic Markov model, age-, sex-, and calendar time-specific transition probabilities (derived from 2002–2012 data of the English Longitudinal Study of Ageing, ELSA) were used to predict prevalence of each health state at the next calendar year. As observed in ELSA, we assumed cardiovascular disease incidence will continue to decline in parallel to cardiovascular mortality. Based on the decline in prevalence of dementia observed in the Rotterdam and Cognitive Function and Ageing Study (CFAS), we inferred a relative reduction of 2% per annum in the incidence of dementia. Validity of the model was established by comparing model estimates for recent years with independent observations.

Results Using ELSA 2006 data as input resulted in predicted age and sex-specific prevalence of dementia for 2011 that was consistent with observed prevalence in CFAS 2011, underpinning the validity of IMPACT-BAM model estimates. As of 2015, we estimated 842,000 people aged 65 or older live with dementia in England and Wales, representing a prevalence of 8.15% (95% CI 3.12–14.24) in this age-group. The number (prevalence) of people aged 65 or older living with dementia is expected to rise to 1,045,000 (9.3%)in 2020, 1,216,000 (9.8%) in 2015, and 1,358,000 (9.7%) in 2030. Prevalence of dementia is and will be higher in women than in men: 9.4% (95% CI 3.0–17.6) vs 6.7%, (95% CI 3.3–10.1), respectively in 2015 and 10.4% (95% CI 2.7–26.3) vs 8.9% (95% CI 4.8–16.5) in 2030.

Conclusion Despite a projected decline in age-specific dementia incidence, the age-specific prevalence of dementia is likely to rise in the next 15 years, largely due to increasing longevity. We predict a substantial increase in the number of people living with dementia, resulting from rising dementia prevalence compounded by population ageing.