Subgroup analysis; Association of primary outcomes at 20 weeks between HG with ongoing vomiting and HG with no ongoing vomiting.

*Score variables are presented as adjusted mean differences (95% CI) from data at 20+ visit, analysed using linear regression with robust estimation.**Edinburgh Postnatal Depression Score categorical variable presented as odds ratio and calculated using ordered logistic regression.Models were adjusted for maternal age, smoking, alcohol, ethnic origin, BMI, marital status, SCOPE centre and job status.N = 5 excluded due to small numbers. These were women with HG before 15 w visit, ceased before 15 w visit, restarted vomiting between 15+ and 20+ visits.