Low economic growth, health, health inequalities and sustainable development goals in a rich country : 27-year Japanese time series

Background: Sustainable Development Goal #8 refers to decent work and economic growth. In the context of climate change and global resource depletion, it is important to understand whether low economic growth is compatible with positive population health in rich countries, particularly because the past decade of austerity in the UK and USA is associated with stagnating life expectancy. Japan provides a natural experiment in that it has experienced low economic growth since 1992, and life expectancy continued to improve. However, the trend in health inequality in good self-rated health is unknown. Methods: We examined trends in health and health inequalities using ten triennial waves of a nationally representative survey in Japan, 1986–2013 (n=731,647). Change in age-standardized self-rated good health was calculated, and health inequalities and their time trends were calculated using Slope and Relative Indices of Inequality (SII and RII respectively) in relation to net household income. Analyses were stratified by sex and age, for children (6–18 years), working-age adults (20–59 years), younger old (60–69 years) and older old (70–79 years), given age differences in relation to the economy and labour market. Time trends of SII and RII were tested during the period of economic stagnation 1992–2013. Results: Overall, age-standardised self-rated good health was high among children (70%) and low among the older old (30%). In all age groups, prevalence of good health declined slightly from its peak in 1995 but increased after 2007. In 1992 among children, working-age adults and younger old, health inequality based on SII for net household income was small (approximately 10% lower prevalence of good health in those with lowest compared to highest income). Among working-age adults, time trends of health inequalities between 1992 and 2013 were curvilinear. The SII narrowed and then widened after 2002 (quadratic trends in men and women p<0.05), resulting in the magnitude o