An analysis of the determinants of anaemia in pregnant women in rural Malawi - A basis for action

Haematological data are presented on 4104 pregnant women attending the antenatal-care facilities of two hospitals in a rural area in southern Malawi. In this area, malaria transmission is perennial and there is a high prevalence of HIV infection. The local women are exposed to drought and food shortages but experience high fertility rates. Mean (S.D.) haemoglobin (Hb) concentration was significantly lower in the primigravidae [8.7 (1.6) g/dl] than in the secundigravidae [9.1 (1.5) g/dl; P < 0.0001] or multigravidae [9.2 (1.5) g/dl; P < 0.0001]. Primigravidae also experienced significantly more iron deficiency and malaria than secundi -or multi-gravidae. For all parity groups, the lowest mean Hb levels were observed between 26-30 weeks' gestation. In primigravidae peak prevalence of malaria occurred between 16-20 weeks' gestation (38.6%) and peak prevalence of moderately severe anaemia (< 8 Hb/dl) between 26-30 weeks' (35.7%). Multigravidae showed little variation in the prevalence of anaemia, iron deficiency and malaria with gestational age. Peak prevalences of malaria were observed in April, in the post-rainy season, with values of 51.4%, 56.0% and 25.3% for primi-, secundi- and multi-gravidae, respectively. Peak prevalences of iron deficiency occurred between April and May and those of moderately severe anaemia between May and June. Mean Hb was lower in adolescent primigravidae than in any other group of pregnant women [8.6 (1.5) g/dl], including the non-adolescent primigravidae [8.9 (1.6) g/dl; P = 0.008]. Other factors significantly associated with increased risk of moderately severe anaemia in primigravidae were illiteracy and poor nutritional status (i.e. body mass index < 18.5 kg/m2 and mid-upper-arm circumference < 23 cm). After forward, step-wise, regression analysis of relative risk (RR) factors and their 95% confidence intervals (CI), variables associated with an increased risk for moderately severe anaemia were iron deficiency (RR = 4.2; CI = 3.0-6.0) and malaria parasitaemia (RR = 1.9; CI = 1.3-2.7) in primigravidae, iron deficiency (RR = 4.1; CI = 2.7-6.3) and mid-upper-arm-circumference < 23 cm (RR 1.8; CI = 1.1-3.0) in secundigravidae, and iron deficiency in multigravidae (RR = 3.1; CI = 4.3-6.9). The basis of anaemia prevention in this population of pregnant women is malaria control and haematinic supplementation, one of the most serious drawbacks being non-compliance. Although the present data are presented according to the World Health Organization's definitions of anaemia, the corresponding cut-off values for Hb (< 11 or < 7 g/dl) were not associated with malaria, suggesting that these Hb levels would be less useful indicators in malaria interventions. It is argued that an Hb value of < 8 g/dl might be considered for identifying the pregnant women at highest risk, for selective health education to reduce non-compliance.