Drs Chandra and Sudhir appear to have misunderstood the aim of our paper. The paper is not primarily about post-partum depression. We did not aim to study the overall aetiology of post-partum depression nor did we aim to accurately predict post-partum depression from the negative self-schema measure. We did aim to test an important aspect of the cognitive theory of depression, namely whether a measure of negative self-schema is associated with the onset of depression. We found there was an association and that it was equally strong whether the onset was during pregnancy, in the post-partum period or 3 years later. In the main analysis we adjusted for the psychological and socio-economic variables outlined in Table 3.

We agree, as stated in our discussion, that a more detailed questionnaire such as the Dysfunctional Attitude Scale would have provided a more comprehensive measure of self-schema. Furthermore, repeated measures would have allowed comparison with other studies and a test of the stability of these ‘schemas’. It is possible that schemas change with an experience such as motherhood, although theoretically they should be relatively stable. As these were secondary data analyses of an existing data-set, we were limited to the data available to us and these did not include any more-detailed or repeated measures of schemas.

There are clearly multiple factors that influence the onset of depression. The correspondents ask why some women have earlier onset than others. This may well be related to changing support or adverse events, but it was not the aim of our paper to address this question. Rather than speculate, the ALSPAC data-set provides an opportunity to answer this question and many others by undertaking further detailed analyses of those data.

The analyses we presented in Table 4 indicate that the strength of the association between negative self-schema and onset of depression does not diminish with time, so it is unlikely that those in the highest tertiles for depression have onset which is sooner.

This work needs to be replicated in other large longitudinal studies of both non-pregnant women and men.