Traditional massage of newborns in Nepal: implications for trials of improved practice.

Mustard oil massage of newborns is an integral component of traditional care practices in many communities. Recent evidence suggests that this practice may have detrimental effects, particularly for preterm infants or for those whose skin barrier function is otherwise sub-optimal. Other natural oils such as sunflower, sesame or safflower seed oil may have a beneficial impact on newborn health and survival. Little is known, however, about cultural and other factors related to the acceptance and uptake of alternative, more beneficial oils for massage of the newborn. A questionnaire concerning the usage and reasons for application of mustard and other oils to newborn skin was administered to the caretakers of 8580 newborns in Sarlahi district of rural Nepal. Four focus group discussions among representative groups were conducted to describe the perceived benefits of oil massage and the factors involved in the decision to apply oil. The potential for the introduction of alternative natural oils was explored. Approximately 99 per cent of newborns were massaged at least once with mustard oil in the 2 weeks after birth, and 80 per cent were massaged at least twice daily. Promotion of strength, maintenance of health, and provision of warmth were the most commonly cited reasons for application of mustard oil. Focus group discussion participants noted that smell, oiliness, mode of pre-massage preparation, and perceived absorptive potential on the skin are important contextual factors involved in the practice. Caretakers are willing to consider adaptation of established traditions for the promotion of positive health outcomes if essential contextual criteria are met. An understanding of cultural, social, and economic factors that shape the context of traditional healthcare practices is essential to the design and implementation of intervention trials examining the relative efficacy of application of oils in reducing neonatal mortality and morbidity.