How do 'effects' differ across time? Understanding health inequalities by triangulating across multiple data sources and empirical strategies

Socioeconomic circumstances such as our education or income are thought to have important effects on our health. On average, the more socioeconomically advantaged someone is, the better their health. Contrary to widely-held myth, these differences are not inevitable or unchangeable. Economic and health policies change across time, and are ultimately expected to either worsen (widen) or improve (narrow) these differences. Scientists, government departments, and voluntary organisations all agree that inequalities in health should be reduced. In order to do so, we need to have an evidence-based understanding of how these inequalities have changed across time to help us develop effective policies. This project seeks to generate this evidence. First, it will bring together multiple datasets to investigate how these health inequalities have changed across time in the UK. In the past, researchers have tended to use these studies separately - analysing them together is a very powerful way of understanding how inequalities have changed across time.

It will then seek to understand whether these correlations reflect causal relationships. While it is possible that social circumstances are correlated with health, there are other explanations for such correlations. It is also possible that such causal effects differ across time, as society changes; and across age, as people get older. It will do this by using multiple data sources which follow many thousands of individuals across time. It will use information on the environment along with genetic information to attempt to address this.

In addition, this project will investigate how the causal role of two other factors may have changed across time. The first is cognitive capability - this refers to how individuals think and process information. On average, those with higher cognitive capability seem to have better health across a range of health measures such as lower cardiovascular disease risk and lower body weight. Due to the increasingly complex nature of society (and potential future changes related to technological development), it is important to understand if the importance of this factor for health has changed across time. The second factor is genetic propensity to ill health. In recent decades our understanding of genetics has greatly improved. We all contain genetic propensity to common health measures, such as body weight and blood pressure, but some individuals have higher genetic propensity than others. These differences do not mean that they will definitively have worse health measures - genetic risk is considered 'probabilistic' rather than 'deterministic'. The expression of these genes also depends on the environment. As such, this project will seek to understand if the links between genetic factors and body weight, blood pressure and mental health have changed across time.

Taken together, this project seeks to advance scientific understanding and provide evidence which can ultimately be used to inform national policies related to public health inequalities in the UK.

The recently purported causal and DNA 'revolutions' in science have largely focused on establishing the presence or absence of causal effects; yet causal effects related to societal factors are likely to differ across time. This project seeks to address this, by evaluating how the effects of socioeconomic, cognitive and genetic factors on health have changed across time - both year of birth (1946 to 2001) and age (maximal span: infancy to old age). This will also provide evidence on key unresolved issues relating to causality in health inequalities.

This is a secondary data analysis project, consisting of the below objectives:

Objective 1: Evaluate whether the effects of socioeconomic position on health outcomes have changed across time.

Objective 2: Evaluate whether the effects of cognition capability on health outcomes have changed across time.

Objective 3: Evaluate whether the effects of common genetic variants on health outcomes have changed across time.

The empirical approach will triangulate across multiple data sources and analytical approaches in order to provide robust evidence. Multiple cross-sectional and longitudinal datasets will first be used to examine change across time in socioeconomic inequalities in health outcomes (body mass index, blood pressure, and mental health - depression and anxiety spectra); this will advance knowledge of descriptive change in health inequality. A range of longitudinal studies will then be used to investigate causal effects of socioeconomic, cognitive and genetic factors on health outcomes. These include the UK birth cohort studies, initiated in 1946, 1958, 1970, 1991, and 2001, and a large Norwegian study (HUNT). Multiple approaches to causal estimation will be used - using observed life course data, and genetic data (including use of Mendelian Randomisation). This will provide new knowledge on how causal effects may change across time.

Policymakers
Governmental and third sector organisations may benefit from this research, since health inequalities are important policy concerns nationally and globally. Reducing health inequalities remains an important arguably unachieved policy goal, with varied policy agendas across time, and ongoing health inequality monitoring initiatives - these largely focus on adult socioeconomic circumstances (area of residence) and binary health outcomes.

This project seeks to better understand how health inequalities have changed across time - by using multiple different sources of data, different indicators of socioeconomic circumstances across life, and examining both average and distributional differences in continuous health outcomes (body mass index, blood pressure, and mental health spectra). This will benefit policymakers by providing more robust evidence on how socioeconomic inequalities in health have changed across time and by indicating when in life may be most optimal to intervene. Consideration distributional differences in health may also inform the types of interventions which are required - if health inequalities are found across the distribution of health, below and above treatment thresholds, then targeting treatment alone will not fully ameliorate health inequalities.

Providing more robust evidence on whether socioeconomic inequalities in health have narrowed, widened or remained stable across time is important to help understand the effectiveness of previous policies. It will provide evidence to examine whether the total effect of previous policy agendas have been sufficient to reduce health inequalities.

This project will also provide more evidence on whether health inequalities, and their change across time, are causal in nature. While most policy-impacting evidence on health inequalities are descriptive in nature, in the longer term understanding the specific causal factors which affect health is likely to be important to ensure that those factors are successfully targeted to improve health.

Additionally, this project will also provide new evidence the changing importance of cognitive and genetic risk factors for health. These factors may be important to consider - alongside socioeconomic factors - in future public health monitoring and preventative health efforts. Indeed, there is currently no consensus on what health differences are 'inequities' or require future monitoring.

While primarily of likely interest to public health-related organisations - for example Public Health England, The Department of Health, King's Fund, Health Foundation - it may also benefit policy-makers working primarily in non-health related fields, since the project will provide evidence on the determinants of health which are related to the broader social and economic environment, and health inequalities have large estimated economic and non-economic social costs. These may include organisations such as the Departments of Work and Pensions, Department of Education, Joseph Rowntree Foundation, the Equality Trust.

The public
The general public may benefit from this research through an improved understanding of the environmental and genetic determinants of health. The increasing popularity of direct-to consumer genetic tests suggest that there is considerable public interest in understanding the links between genetic propensities and important life outcomes. Adding to this, a recently announced Government initiative suggested that future new-borns may genome-sequenced. There is also considerable public interest across the political spectra in the role of socioeconomic and cognitive factors on health.

Commercial sector
Finally, the project will provide new evidence on how the relative importance of genetic variance may differ across time (both age and year of birth) - this may be relevant information to consider for companies which provide information on genetic propensities to consumers.