Background The developmental model of the origins of chronic diseases has shown the importance of undernutrition and poor development in utero on disease in later life, including coronary heart disease and stroke. Tracing disease back to fetal development means that this period can also be seen as the origin of health inequalities and, as such, a priority time for intervention to improve health outcomes. This paper reports on the evaluation of Health in Pregnancy (HiP) grants, a UK-wide unconditional universal cash transfer of £190 made to pregnant women who had sought medical advice before the 25^th week of pregnancy. The grant was designed to provide additional financial support towards a healthy lifestyle including diet, with the cash element providing a greater incentive to seek advice at the appropriate time.

Methods This natural experiment was evaluated using routinely collected data for all singleton births in Scotland before (2004–2009), during (2009–2011) and after (2011–2013) the introduction of the HiP grants. In addition to the primary outcome of birthweight (BW) we evaluated the effect on other measures of stage and size and maternal behaviour including contacting the midwife before 25 weeks. We looked for evidence of differential effects among subgroups including those defined by area deprivation, social class, marital status and maternal diabetes. The analysis was restricted to Scotland because of the high quality routine data available. We used interrupted time series models adjusted for time trends and seasonality in addition to maternal and birth characteristics to estimate the effect of the intervention (INT) and post-intervention (POST) periods relative to the years pre-intervention, and used multiple imputation to reduce any bias due to missing data.

Results Among the 525,400 singleton births from 2004–2013 there was no significant relationship with BW (INT = −2.3 g, 95% CI: −1.9 to 6.6) or other measures of size and stage. The intervention was associated with an increase in the odds of booking before 25 weeks that disappeared post-intervention (INT OR = 1.10, 95% CI: 1.02 to 1.18; POST OR = 0.91, 95% CI: 0.83–1.00). Results for the subgroups largely mirrored those seen for the population.

Conclusion The small financial incentive did not have an impact on birthweight or other measures of size or stage but did appear to have changed health seeking behaviour. It is not clear that a threshold of 25 weeks is optimal to improve birth outcomes given that this is late in the pregnancy. Future evaluations would be enhanced if some randomisation were included in the study design.