Tony Blakely

Unemployment and suicide. Evidence for a causal association

Objectives: To determine the independent associations of labour force status and socioeconomic position with death by suicide. Design: Cohort study assembled by anonymous and probabilistic record linkage of census and mortality records. Participants: 2.04 million respondents to the New Zealand 1991 census aged 18–64 years. Main outcome measure: Suicide in the three years after census night. Results: The age adjusted odds ratios (95% confidence intervals) of death by suicide among 25 to 64 year olds who were unemployed compared with employed were 2.46 (1.10 to 5.49) for women and 2.63 (1.87 to 3.70) for men. Similarly increased odds ratios were observed for the non-active labour force compared with the employed. Strong age only adjusted associations of suicide death with the socioeconomic factors of education (men only), car access, and household income were observed. Compared with those who were married on census night, the non-married had odds ratios of suicide of 1.81 (1.22 to 2.69) for women and 2.08 (1.66 to 2.61) for men. In a multivariable model the association of socioeconomic factors with suicide reduced to the null. However, marital status and labour force status remained strong predictors of suicide death. Unemployment was also strongly associated with suicide death among 18–24 year old men. Sensitivity analyses suggested that confounding by mental illness might explain about half, but not all, of the association between unemployment and suicide. Conclusions: Being unemployed was associated with a twofold to threefold increased relative risk of death by suicide, compared with being employed. About half of this association might be attributable to confounding by mental illness.

Ecological effects in multi-level studies

Multi-level research that attempts to describe ecological effects in themselves (for example, the effect on individual health from living in deprived communities), while also including individual level effects (for example, the effect of personal socioeconomic disadvantage), is now prominent in research on the socioeconomic determinants of health and disease. Such research often involves the application of advanced statistical multi-level methods. It is hypothesised that such research is at risk of reaching beyond an epidemiological understanding of what constitutes an ecological effect, and what sources of error may be influencing any observed ecological effect. This paper aims to present such an epidemiological understanding. Three basic types of ecological effect are described: a direct cross level effect (for example, living in a deprived community directly affects individual personal health), cross level effect modification (for example, living in a deprived community modifies the effect of individual socioeconomic status on individual health), and an indirect cross level effect (for example, living in a deprived community increases the risk of smoking, which in turn affects individual health). Sources of error and weaknesses in study design that may affect estimates of ecological effects include: a lack of variation in the ecological exposure (and health outcome) in the available data; not allowing for intraclass correlation; selection bias; confounding at both the ecological and individual level; misclassification of variables; misclassification of units of analysis and assignment of individuals to those units; model mis-specification; and multicollinearity. Identification of ecological effects requires the minimisation of these sources of error, and a study design that captures sufficient variation in the ecological exposure of interest.

Inequalities in non-communicable diseases and effective responses.

In most countries, people who have a low socioeconomic status and those who live in poor or marginalised communities have a higher risk of dying from non-communicable diseases (NCDs) than do more advantaged groups and communities. Smoking rates, blood pressure, and several other NCD risk factors are often higher in groups with low socioeconomic status than in those with high socioeconomic status; the social gradient also depends on the countrys stage of economic development, cultural factors, and social and health policies. Social inequalities in risk factors account for more than half of inequalities in major NCDs, especially for cardiovascular diseases and lung cancer. People in low-income countries and those with low socioeconomic status also have worse access to health care for timely diagnosis and treatment of NCDs than do those in high-income countries or those with higher socioeconomic status. Reduction of NCDs in disadvantaged groups is necessary to achieve substantial decreases in the total NCD burden, making them mutually reinforcing priorities. Effective actions to reduce NCD inequalities include equitable early childhood development programmes and education; removal of barriers to secure employment in disadvantaged groups; comprehensive strategies for tobacco and alcohol control and for dietary salt reduction that target low socioeconomic status groups; universal, financially and physically accessible, high-quality primary care for delivery of preventive interventions and for early detection and treatment of NCDs; and universal insurance and other mechanisms to remove financial barriers to health care.

Metropolitan area income inequality and self-rated health—a multi-level study

We examined the association of income inequality measured at the metropolitan area (MA) and county levels with individual self-rated health. Individual-level data were drawn from 259,762 respondents to the March Current Population Survey in 1996 and 1998. Income inequality and average income were calculated from 1990 census data, the former using Gini coefficients. Multi-level logistic regression models were used. Controlling for sex, age, race, and individual-level household income, respondents living in high, medium-high, and medium-low income inequality MAs had odds ratios of fair/poor self-rated health of 1.20 (95% confidence interval 1.04-1.38), 1.07 (0.95-1.21), and 1.02 (0.91-1.15), respectively, compared to people living in the MAs with the lowest income inequality. However, we found only a small association of MA-level income inequality with fair/poor health when controlling further for average MA household income: odds ratios were 1.10 (0.95-1.28), 1.01 (0.89-1.14), and 1.00 (0.89-1.12), respectively. Likewise, we found only a small association of county-level income inequality with self-rated health although only 40.7% of the sample had an identified county on CPS data. Regarding the association of state-level income inequality with fair/poor health, we found the association to be considerably stronger among non-metropolitan (i.e. rural) compared to metropolitan residents.

Effects of price discounts and tailored nutrition education on supermarket purchases: a randomized controlled trial

BACKGROUND Traditional methods to improve population diets have largely relied on individual responsibility, but there is growing interest in structural interventions such as pricing policies. OBJECTIVE The aim was to evaluate the effect of price discounts and tailored nutrition education on supermarket food and nutrient purchases. DESIGN A 2 x 2 factorial randomized controlled trial was conducted in 8 New Zealand supermarkets. A total of 1104 shoppers were randomly assigned to 1 of the following 4 interventions that were delivered over 6 mo: price discounts (12.5%) on healthier foods, tailored nutrition education, discounts plus education, or control (no intervention). The primary outcome was change in saturated fat purchased at 6 mo. Secondary outcomes were changes in other nutrients and foods purchased at 6 and 12 mo. Outcomes were assessed by using electronic scanner sales data. RESULTS At 6 mo, the difference in saturated fat purchased for price discounts on healthier foods compared with that purchased for no discount on healthier foods was -0.02% (95% CI: -0.40%, 0.36%; P = 0.91). The corresponding difference for tailored nutrition education compared with that for no education was -0.09% (95% CI: -0.47%, 0.30%; P = 0.66). However, those subjects who were randomly assigned to receive price discounts bought significantly more predefined healthier foods at 6 mo (11% more; mean difference: 0.79 kg/wk; 95% CI: 0.43, 1.16; P < 0.001) and 12 mo (5% more; mean difference: 0.38 kg/wk; 95% CI: 0.01, 0.76; P = 0.045). Education had no effect on food purchases. CONCLUSIONS Neither price discounts nor tailored nutrition education had a significant effect on nutrients purchased. However, the significant and sustained effect of discounts on food purchases suggests that pricing strategies hold promise as a means to improve population diets.

The contextual effects of neighbourhood access to supermarkets and convenience stores on individual fruit and vegetable consumption

Background: It is often suggested that neighbourhood access to food retailers affects the dietary patterns of local residents, but this hypothesis has not been adequately researched. We examine the association between neighbourhood accessibility to supermarkets and convenience stores and individuals’ consumption of fruit and vegetables in New Zealand. Methods: Using geographical information systems, travel times from the population-weighted centroid of each neighbourhood to the closest supermarket and convenience store were calculated for 38 350 neighbourhoods. These neighbourhood measures of accessibility were appended to the 2002–3 New Zealand Health Survey of 12 529 adults. Results: The consumption of the recommended daily intake of fruit was not associated with living in a neighbourhood with better access to supermarkets or convenience stores. Similarly, access to supermarkets was not related to vegetable intake. However, individuals in the quartile of neighbourhoods with the best access to convenience stores had 25% (OR 0.75, 95% CI 0.60% to 0.93%) lower odds of eating the recommended vegetable intake compared to individuals in the base category (worst access). Conclusion: This study found little evidence that poor locational access to food retail provision is associated with lower fruit and vegetable consumption. However, before rejecting the commonsense notion that neighbourhood access to fruit and vegetables affects personal consumption, research that measures fruit and vegetable access more precisely and directly is required.

Food Pricing Strategies, Population Diets, and Non-Communicable Disease: A Systematic Review of Simulation Studies

A systematic review of simulation studies conducted by Helen Eyles and colleagues examines the association between food pricing strategies and food consumption and health and disease outcomes.

Measuring cancer survival in populations: relative survival vs cancer-specific survival

BACKGROUND Two main methods of quantifying cancer patient survival are generally used: cancer-specific survival and relative survival. Both techniques are used to estimate survival in a single population, or to estimate differences in survival between populations. Arguments have been made that the relative survival approach is the only valid choice for population-based cancer survival studies because cancer-specific survival estimates may be invalid if there is misclassification of the cause of death. However, there has been little discussion, or evidence, as to how strong such biases may be, or of the potential biases that may result using relative survival techniques, particularly bias arising from the requirement for an external comparison group. METHODS In this article we investigate the assumptions underlying both methods of survival analysis. We provide simulations relating to the impact of misclassification of death and non-comparability of expected survival for cause-specific and relative survival approaches, respectively. RESULTS For cause-specific analyses, bias through misclassification of cause of death resulted in error in descriptive analyses particularly of cancers with moderate or poor survival, but had smaller impact in analyses involving group comparisons. Relative survival ratio (RSR) estimations were robust in relation to non-comparability of comparison populations for single RSR but were less so in group comparisons where there was large variation in survival. CONCLUSIONS Both cause-specific survival and relative survival are potentially valid epidemiological methods in population-based cancer survival studies, and the choice of method is dependent on the likely magnitude and direction of the biases in the specific analyses to be conducted.

Neighbourhood access to open spaces and the physical activity of residents: A national study

OBJECTIVE Increasing population levels of physical activity is high on the health agenda in many countries. There is some evidence that neighbourhood access to public open space can increase physical activity by providing easier and more direct access to opportunities for exercise. This national study examines the relationship between travel time access to parks and beaches, BMI and physical activity in New Zealand neighbourhoods. METHODS Access to parks and beaches, measured in minutes taken by a car, was calculated for 38,350 neighbourhoods nationally using Geographic Information Systems. Multilevel regression analyses were used to establish the significance of access to these recreational amenities as a predictor of BMI, and levels of physical activity and sedentary behaviour in the 12,529 participants, living in 1178 neighbourhoods, of the New Zealand Health Survey 2002/3. RESULTS Neighbourhood access to parks was not associated with BMI, sedentary behaviour or physical activity, after controlling for individual-level socio-economic variables, and neighbourhood-level deprivation and urban/rural status. There was some evidence of a relationship between beach access and BMI and physical activity in the expected direction. CONCLUSIONS This study found little evidence of an association between locational access to open spaces and physical activity.

What is the contribution of smoking and socioeconomic position to ethnic inequalities in mortality in New Zealand

BACKGROUND Mortality rates for Māori are twice those for non-Māori in New Zealand. We have assessed the contribution of tobacco smoking and socioeconomic position to these inequalities in 45-74-year-old census respondents during 1981-84 and 1996-99 (2.3 and 2.7 million person-years, respectively). METHODS We used linked census and mortality cohort datasets with measures of socioeconomic position (household income, highest educational qualification, car access, labour-force status, and neighbourhood deprivation) and smoking (never, ex, current). We used direct standardisation to adjust for smoking and Poisson regression to adjust for socioeconomic position. FINDINGS The apparent contribution of smoking to mortality differences between Māori and non-Māori non-Pacific people was greatest for women in 1996-99 (8% reduction in standardised rate difference); it had increased from 3% in 1981-84. The corresponding reductions in men were 5% in 1996-99 and -1% in 1981-84. The apparent contribution of socioeconomic factors to mortality differences between Māori and non-Māori non-Pacific was greatest for men (39% in 1981-84 and 37% in 1996-99) and increased over time for women (from 23% in 1981-84 to 32% in 1996-99). INTERPRETATION Although small, the contribution of smoking to ethnic inequalities in mortality increased over time and might grow more during the next two decades if differences in smoking between ethnic groups continue to increase. Better measurement of socioeconomic position (eg, lifecourse measures, asset wealth) might increase the proportion of ethnic inequalities attributable to socioeconomic position, perhaps to about half. Action to redress socioeconomic gaps and control of tobacco use will both be important in reducing ethnic inequalities in health.