Table 1.

Mechanistic assumptions and their application to social determinants of health^a

Assumption	Description	How the assumption circumvents complexity of health inequities
Reductionism	Assumes the whole system can be understood by identifying, describing and analysing all of its constituent parts	Breaks social determinants of health into separate, distinct factors (rather than a set of complex intersecting factors) Draws attention to symptoms or expressions of root causes that are more immediately visible (e.g. considering ‘race’ a determinant of health instead of ‘racism’)
Linearity	Assumes that (1) output changes proportionally with input, and (2) the effect of combined inputs can be understood and predicted by dissecting the input–output relationships of individual components, or a direct summative and predictive cumulation of constituent parts	Simplifies interconnectedness Justifies use of proxy indicators that reflect symptoms rather than causes of health inequities (e.g. monitoring maternal and child mortality rates as indicators of equity)
Hierarchy	Assumes central power and control, which diffuses systematically from proximal to more distal parts	Places responsibility for acting on health within individuals or groups, rather than society Legitimises a focus on health damaging behaviours rather than health damaging conditions, systems or structures

^aAdapted from Jayasinghe, 2011 [20]