Malaria is an important environmental factor which reduces fetal growth in primiparae more than multiparae living under holoendemic conditions for malaria. This relates to greater susceptibility to malaria infection in first pregnancies. The relative risk for low birthweight (less than 2500 g) associated with primiparity is increased in malaria-endemic areas and significantly correlates with the malaria parasite rate at delivery in primiparae. Because of this association, the relative risk is proposed as an indicator to assess malaria control in pregnant women as well as in the community. The sensitivity and specificity of the relative risk for low birthweight in primiparae are calculated for 13 malarious and 15 non-malarious populations. The highest sensitivity and specificity is achieved at a relative risk of 1.7. Social and environmental variables which could alter the sensitivity of the estimate are discussed. Estimates of the population-attributable risk per cent of low birthweight due to malaria in primiparae are calculated and vary between 10% and 40% in endemic areas. The method is applied to observations from malaria-intervention studies in pregnancy in the Solomon Islands and Papua New Guinea and appears sensitive in these prospective studies to changes in malaria prevalence. Calculation of these estimates is straightforward and their use to assess malaria control measures in areas of high transmission has not been suggested previously, it could have wide epidemiological application and requires further field evaluation.
PIP: The risk ratio (RR) was used to determine the population attributable risk (PAR) percent for low birth weight (LBW) primiparous women to evaluate malaria control in pregnancy and in community in developing countries. In holoendemic countries, the RR for LBW is associated with the parasite rate in these women at delivery. Moreover, malaria does not have such a strong effect on fetal growth reduction in multiparous women as it does in primiparous women. The sensitivity for 13 malarious areas and specificity for 15 nonmalarious areas was also determined. The highest sensitivity and specificity was found at an RR of 1.7. A longitudinal study in rural Madang, Papua New Guinea shows that despite pregnant women adequately taking their weekly chloroquine prophylaxis provided by the monthly mobile clinics, parasite clearance was poor, indicating a high level of chloroquine resistance. Further it reveals that the RRs did not decrease with increased antenatal visits and incidence rates of Plasmodium falciparum stayed high (20%/person-month). Under these circumstances, the RR was a sensitive indicator of poor malaria control. Other research shows that the RR for LBW in rural Gambia corresponded with that for coastal Papua New Guinea (2.9) which indicates that malaria affects women experiencing their 1st pregnancy equally without regard to differences in mean birth size between the 2 populations. A study reports the changes in the percentage of LBW infants delivered during a malaria control program (spraying to control mosquitoes) in the Solomon Islands. The reduction in the PAR percent for LBW during malaria control in the study population was estimated to be 24.1% and the actual reduction in LBW among primiparous women was 22.1%. Therefore the RR and the PAR percent are sensitive indicators of malaria control. Researchers should report parity-specific LBW percentages thereby enabling them to calculate the RR.