c Reviews mate choice by males? Aiibn. Beliuv. 40,870-876 sitaires de Bordeaux 42 (1979) 47 (1994) Burley, N. and Moran, N. The significance of age and De Clercq, D. et nl. Schistosomiasis in Dogon country, reproductive experience in the preferences of feral pigeons, Mali: identification and prevalence of the species responsible 27, 686-698 Coliciiibn livia. Aiziiii. Belinzi. for infection in the local community. Trniis. R. Soc. Trop. Med. 43 (1994) 88,653-656 Owens, I.P.F. et nl. Sex differences, sex ratios and sex HI&. roles. Proc. R. Soc. Loildoti Ser. B 258,93-99 48 Woolhouse, M.E.J. (1991)On the application of mathematical 44 (1987) 116,80-88 Combes, C. et nl. Les schistosomes.Porir In Sckiice models of schistosome transmission dynamics. I. Natural 45 Southgate, V.R. (1978) On factors possibly restricting the transmission. Actn Trop. 49,241-270 distribution of Scliistosoiiia intercnlatuiii Fisher, 1934. Z. Pnm- 49 May, R.M. and Woolhouse, M.E.J.(1993) Biased sex ratios and siteiikiriide 56, 183-193 .. parasite mating probabilities. Pnrnsitologj 107,287-295 46 (1987) 50 (1990) Doumenge, J.P. et nl. Atlas of tlie Globnl. Distri6utioii of Basch, P.F. Scliistosoiiies: Developmiit, Reprodttctioii, niid Schistosoiiiinsis, pp 41-49. CEGET-CNRS-WHO, Presses Univer- Host Relntioiis, pp 152-153, Oxford University Press Focus Combating Malaria Morbidity and Mortality by Reducing Transmission J-F. Trape and C. Rogier - Jean-François Trape aiid Christophe Rogier present epideini- viduals living all their lives in a given endemic area ological data aizd an nizalysis of the relatioizship between depends on transmission intensity and age. This has trarzsnzission, morbidity ancl mortality from inalaria zvhich marked consequences for the absolute and relative suggest that any intervention aiming to reduce trans- importance of the burden of malaria at a given age7, nzissioii will not, on a long-tem basis, reduce the burden of but also, probably, on tlie immediate and delayed malaria in the majority of epidemiological coiztexts observed evolution of the incidence of malaria morbidity and iiz tropical Africa. mortality after a reduction in transmission. I Malaria control in tropical Africa is principally based Transmission and mortality on the presumptive treatment of fever cases using The results of a large study covering the 500000 anti-malarial drugs. In the past decade, the rapid inhabitants of Brazzaville (Congo) provide an initial spread of chloroquine resistance has stimulated the indication that extreme differences in malaria trans- exploration of other control methods. Several studies mission may be associated with only minor differ- have now shown that insecticide-impregnated bed- ences in malaria mortality rates8-11. This study is the nets can reduce morbidity and mortalityl-4, and this only published comparison of malaria mortality rates method is generally considered to be an efficient between populations that were identical in their gen- means of combatting malaria. Aided by substantial etic and socio-cultural backgrounds and that bene- funding from several international agencies, interven- fited from equal opportunities for therapeutic care, tion programmes based on insecticide-impregnated while differing dramatically in their exposure to ma- bednets and curtains are either under way, or being laria. Depending on the district of Brazzaville, the planned, in many African countries. Other strategies entomological inoculation rate varied from more than aimed at reducing malaria transmission (such as the 100 infective bites per person per year to less than one genetic manipulation of mosquito vectors5 or the de- infective bite per person every three years, which rep- velopment of a transmission-blocking vaccine9 are resents almost the entire scale of malaria transmission also actively being explored. rates observed in Africag. Despite this, the incidence Generally speaking, can we hope that interven- of severe malaria cases was essentially identical for all tions that aim to reduce malaria transmission can re- the districts, the only significant difference being the zi duce, on a long-term basis, malaria morbidity and younger average age of severe malaria attacks in the0 P mortality whatever the epidemiological context? The high-transmission districts (Fig. 1). It is important to + .a answer lies within the general framework of relation- note that the parasite rate in schoolchildren varied ships between the entomological inoculation rate, the from 3% to 81%, depending on the district of the0o= a incidence rate of malaria attacks and the frequency of town, and that two-thirds of the schoolchildren from .f! severe forms of the disease. The average level of the low-transmission districts had no detectable anti- transmission varies considerably with the endemic Plasnzodiunz antibodies at the age of seven, which C - area, from about to lo3 infective bites per person clearly indicated that the circulation of children be- a, -7-1 lper year. The degree of acquired iminun6y i; indi- tween different districts was limited and could not, E 2- 0- 0- $- therefore, explain the homogeneity of the risk of se- 2 g-g vere malaria". Recently, other studies have compared c) o =======CI Jean-FrançoisTrape is at the Labot-atoirede Paludologie, 'I severe malaria rates in areas with different entomo- Cl Ic=b-~ ORSTOM, BP 1386, Dakar, Sénégal. Christophe Rogier is u-0 .. E7_1 -0 I logical inoculation rates. Similar levels of severe $ "====Oat the Service d'Epidémiologie, Institut Pasteur, BP 220, Dakar, O- O- -? Sénégal. Tel: +22I 32 O9 62, Fax: +22I 32 I6 75, ? malaria were observed in two areas of markedly c U i c- c- m-0= e-mail: [email protected] different malaria transmission in East Africa: one in ,f Y- Y- 0- O VOI. 12, no. - 236 Copyright 1996. Elsevier Science Ltd All rights reserved O1 69-4758/961 6, I996 - '$1500 PII SOl69-4758(96)lOOl5-6 Parasito/ogyToday, , . .. F~OCUS Tanzania, where transmission 5 reached 300 infective bites per per- .-cL1 son per year; the other in Kenya, 2 where transmission was about 10 24 infective bites per person per year12. As in the case of Brazzaville, the E $3 only important differences con- v)a, cerned the age distribution and the c clinical patterns of the severe forms a,02 o of the disease. Similarly, the inci- C dence of severe malaria was not a, 21o associated with transmission level at -C nine different sites in the Kilifi Dis- O 01 trict in Kenya13. Finally, comparison 01.3-2 2-12 12-52 52-11 of malaria mortality rates observed in 28 studies in Africa revealed that Entomological inoculation rate (EIR) these estimated rates were generally Fig. I. Incidence of severe malaria as the number of cases per 10000 people per year of of only limited variability and observation in children aged 0-14 years living in Brazzaville (Congo) according to the showed no relationship with the entomological inoculation rate (EIR) in the district of residence. Cases occurring before transmission level in the 11 studies age five are shown in black, and those occurring after age five are hatched. where entomological data were available14. ence, night and day, of a medical team in the village to diagnose and treat any pathological episode. As Transmission and morbiditv shown in Fig. 2, these two populations differ markedly If the level of transmissioh is not an important risk in terms of the pattern of age-dependent variations in factor for malaria mortality in Africa, is the same true malaria attack incidence rates. From these data, it can for malaria morbidity? In highly malaria-endemic be estimated that, at the age of 60, the Dielmo inhab- areas, distinguishing malaria from other causes of itants, who are exposed to about 200 infective bites fever poses difficult methodological problems be- per person per year, average a total of 43 attacks since cause of the high frequency of asymptomatic infec- birth, with only 23% of these arising during adult- tions and the lack of specificity of the signs and symp- hood. For the Ndiop inhabitants, who are exposed to toms of the disease's. It is only recently that methods about 20 infective bites per person per year, one can have been developed that permit a precise estimation estimate an average total of 62 malaria attacks by the of the incidence of malaria attacks in areas of moder- age of 60, of which 41 % occur during adulthood. ate and high transmission15-18. By using these The comparison of these three Senegalese popu- methods, we have compared the malaria morbidity of lations suggests that there is little difference in the total three Senegalese populations (from Dakar, Ndiop and number of attacks over an entire lifetime in individ- Dielmo) exposed to approximately 1, 20 and 200 uals residing in areas that vary by as much as a factor infective bites per person per year, respectively. of 200 in transmission intensity. If this is correct, how In Dakar, among individuals that have lived since do we explain the decrease in malaria morbidity and birth in a district of the town where the transmission mortality following the implementation of bednets intensity was about one infective bite per person per programmes which has been observed in most stud- year, the clinical incidence rate was identical to the ies in Africa? Figure 3 compares the fluctuations of parasitological incidence rate in children aged seven the entomological inoculation rate and the incidence to 11 (Ref. 7) and was three times less than the para- density of malaria attacks in chidren in Dielmo. sitological incidence rate in adults (T-F. Trape and Clearly, these fluctuations are closely correlated, and L. Konate, unpublished). These observations show that a tenfold decrease or increase in malaria transmission a high proportion of infections are symptomatic in is associated, in the following weeks, with a twofold individuals of all ages exposed since birth to about decrease or increase in malaria morbidity.