Authors' reply

We investigated the extent to which adult social adversity and ill health contributed to an elevated risk for depressed mood among adults with mild learning disability (Collinshaw et al, 2004). The study used data from the 1958 National Child Development Study (NCDS), a nationally representative cohort followed from birth to age 43 years.

Dr Feroz-Nainar makes the point that epilepsy, fragile-X syndrome and Down's syndrome are among the biological/genetic causes and correlates of learning disabilities and raises the question whether these factors contributed to the higher rate of depressed affect associated with mild learning disability.

A previous report on the NCDS birth cohort confirms that epilepsy and other neurological abnormalities were indeed more common for individuals with mild learning disabilities than for controls. However, the majority of individuals with mild learning disability had no known neuro-epileptic abnormalities and mild learning disability was more commonly associated with childhood social and family adversity (Maughan et al, 1999).

To investigate the possibility that group differences in depressed affect were due to biological factors such as epilepsy in those with mild learning disabilities, we re-analysed the statistical models reported in our recent paper (Collishaw et al, 2004). Controlling for childhood epilepsy/neurological problems did not reduce group differences in adult depressed affect (model adjusted only for gender: OR=2.84, 95% CI 1.7-4.9, P<0.001; model adjusted for gender and childhood neurological problems/epilepsy: OR 2.79, 95% CI 1.6-4.8, P<0.001). This is in contrast to the partial mediating effect of controlling for childhood social adversity (Maughan et al, 1999; Collishaw et al, 2004) and the almost complete mediating effect of additional controls for adult ill health and adult social adversity (Collishaw et al, 2004).

We cannot rule out completely the possibility that some other unmeasured third factor is confounded with social adversity and could explain our findings. We also acknowledge that specific biological factors may be of particular importance for understanding affective problems in some individuals with mild learning disability. Nevertheless, when assessed in an unselected general population cohort such as the NCDS, social factors and adult health do appear to have an important contribution to depressed mood among people with mild learning disability.