Background and purpose: Although women have a greater lifetime risk of stroke than men, along with other gender differences in stroke, there is lack of reliable data on long-term trends in stroke in women. This paper presents the results of three population-based registers in Auckland (1981-1982, 1991-1992, and 2002-2003) which aimed to determine the trends in the incidence of stroke in New Zealand. This paper reports on the burden of stroke in women across different ethnic groups over the study period from 1981 to 2003.
Methods: Three studies using similar methodology and overlapping case finding methods have been conducted in New Zealand to estimate trends in the incidence and outcome of stroke. The studies are recognised as meeting the 'ideal' criteria for stroke incidence and identify all first-ever and recurrent strokes in residents (aged greater than and equal to 15 years) the population of Auckland, New Zealand in the 12-month periods from 1981-1982, 1991-1992, and 2002-2003. There were totals of 307,578 strokes among women in 1981, 372,642 in 1991, and 470,727 in 2001, according to each Census representing over 50% of the population at each period (51%-54%). The WHO standard world population was used for direct age standardisation of annual rates (per 100,000 population), reported with 95% confidence intervals (CI).
Results: The proportion of women who have experienced a stroke did not change markedly over the study period but there were ethnic differences emerging which were related to population changes within the New Zealand population. Standardised stroke incidence in women was relatively stable across the three study periods (1981-1982), 133 [95% CI 118-151]; (1991-1992), 143 [95% CI 116-241]; and (2002-2003), 124 [95% CI 115-134]). However, a significant decline of 14% (95% CI 2%-29%) in rates of first-ever stroke was found between 1991-1992 and 2002-2003 in women. In contrast to the significant declines in event rates in European women over two decades (Rate ratio 0.84, 95% CI 0.73-0.96), increasing trends in event rates in Pacific women were observed (2.71 95% CI 1.00-7.29). The rate of stroke for Maori women did not change significantly over time. Over half of the women who had a stroke event reported that they had high blood pressure; one in four reported that they had diabetes. The proportion of women who smoked declined over time but increases in body mass index (BMI) indicated weight gains in women over time. These trends were consistent across ethnic groups. Women's survival after 1 month following their stroke has improved by 39% (p<0.0001) over the 20-year period. Favourable changes in early survival were most pronounced in European women.
Conclusions: There was a modest decline in stroke incidence in women (overall and for New Zealand European women in particular) in Auckland over the past 20 years but there also was a trend towards increasing stroke incidence in Maori, Pacific, and Asian women. These divergent trends are likely to be associated with different trends in the prevalence of risk factors in these ethnic populations. Targeted stroke prevention programmes are needed in New Zealand to meet the needs of specific ethnic groups as well as the needs of providing ongoing care and support to women following their stroke.