Inequalities in mortality during and after restructuring of the New Zealand economy: repeated cohort studies

Abstract

Objectives: To determine whether disparities between income and mortality changed during a period of major structural and macroeconomic reform and to estimate the changing contribution of different diseases to these disparities.

Design: Repeated cohort studies.

Data sources: 1981, 1986, 1991, 1996, and 2001 censuses linked to mortality data. Population Total New Zealand population, ages 1-74 years.

Methods: Mortality rates standardised for age and ethnicity were calculated for each census cohort by level of household income. Standardised rate differences and rate ratios, and slope and relative indices of inequality (SII and RII), were calculated to measure disparities on both absolute and relative scales.

Results: All cause mortality rates declined over the 25 year study period in all groups stratified by sex, age, and income, except for 25-44 year olds of both sexes on low incomes among whom there was little change. In all age groups pooled, relative inequalities increased from 1981-4 to 1996-9 (RIIs increased from 1.85 (95% confidence interval 1.67 to 2.04) to 2.54 (2.29 to 2.82) for males and from 1.54 (1.35 to 1.76) to 2.12 (1.88 to 2.39) for females), then stabilised in 2001-4 (RIIs of 2.60 (2.34 to 2.89) and 2.18 (1.93 to 2.45), respectively). Absolute inequalities were stable over time, with a possible fall from 1996-9 to 2001-4. Cardiovascular disease was the major contributor to the observed disparities between income and mortality but decreased in importance from 45% in 1981-4 to 33% in 2001-4 for males and from 50% to 29% for females. The corresponding contribution of cancer increased from 16% to 22% for males and from 12% to 25% for females.

Conclusions: During and after restructuring of the economy disparities in mortality between income groups in New Zealand increased in relative terms (but not in absolute terms), but it is difficult to confidently draw a causal link with structural reforms. The contribution of different causes of death to this inequality changed over time, indicating a need to re-prioritise health policy accordingly.

Fig 1 All cause mortality rates (per 100 000) by three income groups by period for all ages (1-74 years) combined. Rates and 95% confidence intervals are presented in tables A and B on bmj.com. Error bars are 95% confidence intervals

Fig 2 Trends in absolute (slope index of inequality (SII)) and relative (relative index of inequality (RII)) inequalities in all cause mortality by income for all ages (1-74 years) combined. Error bars are 95% confidence intervals

Fig 3 All cause mortality rates (per 100 000) by three income groups by period and age group. Rates and 95% confidence intervals are presented in tables A and B on bmj.com

Fig 4 Cause specific mortality rates (per 100 000) by three income groups by period for all ages (1-74 years). Rates and 95% confidence intervals are presented in table C on bmj.com

Fig 5 Contribution of causes of death to SII for all ages (1-74 years) combined. Total height of bars is all cause mortality SII shown in table 2. Height of each component is cause specific SII shown in table 3. Contribution of suicide for females in 1981-4 is shown below 0 as suicide rates were slightly greater among people on high income. Percentage contribution of each cause of death is shown in table D on bmj.com

Fig 6 Contribution of causes of death to SII for 25-44, 45-64 and 65-74 year olds. Graphs for children and youths could not be shown because of insufficient numbers of deaths in high income groups. Components to left of 0 depict those instances where rates were higher in high income people, thus making an “offset” contribution to the total SII