This supplement is dedicated to the memory of Ramesh ‘Channi’ Kumar, Professor of Perinatal Psychiatry at the Institute of Psychiatry, King's College London, and consultant perinatal psychiatrist at the South London and Maudsley NHS Trust, who died in September 2000.

Channi was a key figure in the development of perinatal psychiatry as a specialty, and his research into the causes, consequences and treatment of postnatal mental illness gained him an international reputation. One of his greatest achievements was in raising awareness — in public, medical and academic arenas — about postnatal mental illness and its impact not only on the women who suffer from it but also on their babies and other family members.

Along with Ian Brockington and James Hamilton, Channi launched the Marcé Society (the international society for research into mental illness associated with childbearing) and became its second president, from 1984 to 1986. During this time, the society became a leading forum for a diverse group of researchers and clinicians involved in the treatment of postnatal psychiatric illness — midwives, health visitors, obstetricians, psychologists and paediatricians, as well as mental health specialists. Over the years, with Channi's encouragement, the society has had a substantial impact on routine clinical practice for mothers and their infants.

Channi was a dedicated clinician with real concern and respect for his patients. In 1981 he was appointed consultant to the newly opened mother and baby unit at the Bethlem Royal Hospital. This innovative development allowed mothers with a mental illness to be admitted with their babies rather than being separated from them. Over the years, researchers at the unit have collected a variety of data to provide a sound basis for future clinical care. The unit has recently been renamed the Channi Kumar Mother and Baby Unit in recognition of his contribution to its history and development.

Channi's multi-disciplinary approach was also evident in the setting up of a liaison service combining obstetric and psychiatric resources to care for women in pregnancy and postnatally, which has become a model of its kind. He encouraged the introduction by King's College Hospital of a specialist midwifery service which aims to provide, as far as possible, continuity of midwifery care for pregnant women with current or pre-existing mental illness.

His scientific work was wide-ranging and included studying the biochemical aspects of post-partum mental disorder (Reference Wieck, Hirst, Marks, Racagni, Brunello and FukudaWieck et al, 1991; Reference McIvor, Davies and WieckMcIvor et al, 1996); the extent to which medication taken by the mother to help her mental condition can be passed through her milk to her baby (Yoshida et al, Reference Yoshida, Smith and Kumar1997a ,Reference Yoshida, Smith and Craggs b ); new methods for treating mothers with postnatal depression (see below); and the impact of mothers' postnatal illness on their children's development (Reference Yoshida, Marks and CraggsYoshida et al, 1999; Reference Hay, Pawlby and SharpHay et al, 2001). In a series of studies he followed up his interest in physiological factors that might be involved in the aetiology of post-partum psychosis, in particular oestrogen-related augmented dopaminergic activity in the brain. His research demonstrated that women with a history of bipolar disorder, who are at high risk of post-partum psychosis, do indeed have elevated dopaminergic activity (assessed by measuring growth hormone secretory response to an apomorphine challenge at 4 days post-partum), but that this is a trait effect only and does not predict post-partum psychotic relapse. In contrast, post-partum depressive relapse, in both bipolar and unipolar depressive illness, is predicted by augmented dopaminergic activity (Kumar et al, Reference Kumar, Marks and Wieck1997, Reference Kumar, Marks and Wieck1998).

These findings were translated into clinical practice in a randomised controlled trial which showed that transdermal oestrogen is an effective treatment for postnatal depression (Reference Gregoire, Kumar and EverittGregoire et al, 1996). His most recent study, which he was completing at the time of his death, was an open clinical trial of the efficacy of transdermal oestrogen in preventing recurrence of affective psychosis after childbirth in women at high risk. The study showed that oestrogen did not prevent relapse, but women receiving the highest starting dosage (800 g per day) needed less medication and had shorter periods of hospital admission than those receiving lower starting dosages (400 g or 200 g per day), suggesting that although oestrogen administration may not prevent recurrence of psychosis, it may potentiate the therapeutic response to antipsychotic drugs (Reference Kumar, Davies and McIvorKumar et al, 2003).

Channi collaborated widely with colleagues around the world, many of whom (including several contributors to this supplement) spent time as visiting researchers or clinicians at the Institute of Psychiatry and Maudsley Hospital, and took back home his enthusiasm and ideas for improving psychiatric services for perinatal women. His research collaborators included colleagues from Portugal (Areias et al, Reference Areias, Kumar and Barros1996a ,Reference Areias, Kumar and Barros b ; Reference Augusto, Kumar and CalheirosAugusto et al, 1996), France (Reference Guedeney, Fermanian and GuelfiGuedeney et al, 2000), Japan (Yoshida et al, Reference Yoshida, Marks and Kibe1997c , Reference Yoshida, Marks and Craggs1999; Okano et al, Reference Okano, Nagata and Hasegawa1998a ,Reference Okano, Nomura and Kumar b ) and Chile (Reference Morales, Marks and KumarMorales et al, 1997) as well as from Italy, Brazil, Saudi Arabia and the USA.

Over his last few years, Channi became increasingly interested in cross-cultural issues in perinatal psychiatry. In 1994 researchers from around the world met at a workshop, organised by Channi, to explore the possibilities of setting up a cross-cultural study. The resulting Transcultural Study of Postnatal Depression was instituted in 1998 and was coordinated by Channi until September 2000. His courtesy, charm and ability to make people feel they were respected and valued by him, as much as his considerable abilities as a clinician and scientist, enabled this diverse group of researchers to work together in mutual respect and understanding and thus bring to fruition the study that he had initiated.