How low can it go? Projecting ischaemic heart disease mortality in New Zealand to 2015

This study aims to identify how ischaemic heart disease (IHD) mortality rates in New Zealand have varied between successive cohorts and time periods. This information is then used to project IHD mortality rates and counts (burdens) out to year 2011-15.

Age/period/cohort models were constructed (5-year periods and 5-year age groups, generating 10-year overlapping cohorts) using both frequentist and Bayesian methods. Data were available from 1956 for the total population and from 1981 for Maori. The projection period was 2001-5 to 2011-15. Uncertainty was quantified as the Bayesian 90% credible interval.

IHD mortality rates for all age by gender groups increased from 1956-60 to peak in 1966-70, then declined by more than 60% to current (1996-2000) levels. However, the decline has been much shallower for Maori. This decline has resulted from increasingly favourable period effects since 1971-75 (less marked for Maori). However, no substantive cohort effects have been seen, at least from the 1891 to the 1951 cohort. Our model suggests that, for the first time, a substantive and unfavourable cohort effect may be emerging among recent birth cohorts.

IHD mortality rates are projected to continue to fall from 2001-05 to 2011-15, albeit more slowly than in the past as the increasing (favourable) period effect is partly offset by an emerging (unfavourable) cohort effect. The result is a relatively small projected decline in absolute IHD mortality burden overall, but an actual increase among Maori.