Testing life history theory in a contemporary African population Meij, J.J.

Hele tekst

(1)Testing life history theory in a contemporary African population Meij, J.J.. Citation Meij, J. J. (2008, February 21). Testing life history theory in a contemporary African population. Retrieved from https://hdl.handle.net/1887/12615 Version:. Corrected Publisher’s Version. License:. Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden. Downloaded from:. https://hdl.handle.net/1887/12615. Note: To cite this publication please use the final published version (if applicable)..

(2) 70.

(3) Chapter 4 Low cost interventions accelerate epidemiologic transition in Upper East Ghana. Accepted for publication in: Transactions of the Royal Society of Tropical Medicine and Hygiene. JJ Meij AJM de Craen J Agana D Plug RGJ Westendorp. 71.

(4) Summary Because. effective. medical. interventions. before. the. 1950s. were. merely. absent, the traditional epidemiologic transition in developing countries was mostly exerted by the introduction of water mains, sewage systems, and personal. hygiene.. Nowadays,. effective. medical. interventions. like. vaccination programs, medication, and vitamin supplements might add to a swift. transition.. (N=18. 850). in. We. registered. mortality. the. Garu-Tempane. among. District,. the. Ghana. research (2002. –. population 2005). and. calculated the expected mortality based on the population structure in 2002. Furthermore. we. compared. the. life. expectancy. of. the. region. with. other. countries depending on their GDP. Mortality in the age group 0-9 years was 8.1 per 1 000 person years and in the age group 10-19 years it was 4.1 per 1000 person years. Cumulative survival probability up to age 20 amounted to 89% and was far lower than expected. Observed and expected mortality in old age were similar. The life expectancy at birth was 59 years and much higher than the GDP of $100 would predict. We conclude that the population is in epidemiologic transition. It shows that an epidemiologic transition can be accelerated with low cost interventions.. 72.

(5) Introduction In the early 1970s, Omran (1971) introduced the epidemiologic transition theory. He defined the epidemiologic transition as a societal development in which the pattern of mortality shifts from mortality in childhood mainly caused by communicable diseases, to mortality in late life, predominantly caused by degenerative diseases. Until the 1930s in developed countries, the epidemiologic transition was almost exclusively driven by the introduction of water mains, sewage disposal systems, and school hygiene, often exerted through medical doctors, the so-called hygienists (Wolleswinkel et al., 1997).. Compared. to. the. so-called. ‘developed’. countries,. the. epidemiologic. transition in less developed countries has started at a later time (Andersson and. Moniruzzaman,. 2005;. Huynen. et. al.,. 2005;. Van. Landingham. and. Hirschman, 2005). Up to this moment, there is still great variability between developing countries in the phase of the epidemiologic transition that has been achieved. Moreover, within the least developed countries there is still large variation in child mortality. For example, in Botswana infant mortality is 119 per 1000 births compared to 214 per 1000 births in Burkina Faso (WHO, 2003). Finally, within the developing countries there are marked regional differences in the transitional status, childhood mortality in rural areas. being. in. general. higher. when. compared. to. urban. areas. (Ghana. Statistical Service, 2004; Hammer et al., 2006).. Nowadays, in less developed areas, there are three principal strategies that may contribute to reducing child mortality. First, the improvements in the nutrition status of the family. Second, there are the classic hygienic measures such as clean water supply, sewage disposal systems, and school hygiene. Third, there are more medically oriented interventions such as vaccination, vitamin-A administration, distribution of oral re-hydration solutes, use of medication, bed nets, education of mothers, and spacing of children (Adjuik et al., 2006; Breiman et al., 2004; Ghana VAST study team, 1993; Moor et al., 2003; Vaugelade et al., 2004). Unlike the classic hygienic measures, it is yet unclear whether medically interventions in itself are able to spark an epidemiologic transition.. The. Ghana. Demographic. Health. Survey. 2003. (DHS). has. reported. on. remarkable low child mortality in the underdeveloped Upper East Region of Ghana (Ghana Statistical Service, 2004). The finding was highly disputed and unexpected, as the region has an estimated per capita income of less than 100 dollar per year and for that reason can be classified amongst the least developed regions. As this result was based upon small household samples, questions have raised on the validity of these data and into the mechanisms. 73.

(6) when an epidemiologic transition is taking place. As part of a population based survey into the determinants of child mortality in the Upper East Region of Ghana, we have studied levels of child and infant mortality from 2002 through 2005 and challenged the results of the demographic survey. Furthermore, we contrasted the mortality data with the environmental and hygienic conditions in the area.. Methods We conducted our study in the Garu-Tempane district, a densely populated agricultural area in the south east of the Upper East of Ghana. Compared with the south of Ghana, the whole Upper East region, and in particular the GaruTempane district is underdeveloped. The district has an estimated per capita income of less than 100 US Dollar per year (IFAD, 2006), while for the whole of Ghana it is about 2,150 US Dollar per year (UNDP, 2003). The area has a semi-Saharan climate with an average temperature of 32º C throughout the year and a yearly rain season from June to August. The research area is approximately. 200. km2. with. an. estimated. population. of. about. 18,500. subjects.. The Garu-Tempane district is inhabited by several tribes, mostly Bimoba (66%), Kusasi (22%), Mamprusi (4%), and Fulani (4%). The majority of people are farmers. The total agricultural process is done by hand and any form of mechanized farming is absent. Around 1975 some water boreholes were introduced to the region, although most of the villages in the research area still depend on rainwater and water from the small stream that floats through the area. Sewage disposal systems are non existent. Illiteracy is over 95% for the population aged 15 years and over. In our remote research area bed-nets are almost absent and those available are not very well kept.. Formal medical care is almost absent, although many traditional healers are active in the area. The nearest hospital is in Bawku at a distance of 50 km. Recently several health clinics run by a medical assistant or a nurse have been set up, but these are not yet fully operational. The first clinic that has physicians on the site is at Bawku, 50 kilometres away. At the time of our first year of research only one medical doctor was working in Bawku, serving approximately 400.000 inhabitants of the Bawku/Garu area.. Mass vaccination campaigns of newborns were introduced in the mid 1980s and were predominantly targeted at children up to 12 years of age (EPI, 2005). Vaccination was aimed at reducing polio, measles, whooping cough, diphtheria, tetanus, and BCG. In 2003 and 2004 mass vaccination programs against polio were carried out. Based on a survey of 600 children in two. 74.

(7) randomly selected villages in our research area, we have estimated that in the area at least 50 % of all children under ten years of age have been vaccinated at least once. This is about 25 % lower than the figure for the – overall more developed - Upper East Region, reported in the Demographic Health Survey 2003 (Ghana Statistical Service, 2004). Over the last ten years, additional other mass treatment programs have been undertaken in which all inhabitants of the area were treated with the antihelminth drug albendazole (Ziem, 2005).. For the present research we started a demographic database that was initially set up in 2002 for parasite monitoring (Storey, 2002). This census was carried out as a joint effort by Dutch and Ghanaian field staff with assistance of local translators. Every compound in the research area was mapped by means of a Global Positioning System (GPS) and received a unique compound number (Garmin, 2001), together with the number of residents, their names, sex, tribe, and estimated ages. To optimize the accuracy of the age estimates, the age. of. each. person. was. observed. by. three. different. co-workers.. After. comparing and discussion, the agreed age was registered. Some 2460 women aged over 25 years underwent a detailed interview on their fertility histories including the number of deliveries and the number of children still alive. From 2003 through 2005, we revisited all registered compounds each year, updated the demographics of the inhabitants, registered the persons who deceased over the last year, and registered the migration patterns.. Statistical analysis The statistical analysis proceeded in various stages. First, we constructed an abridged life table based on the observed number of deaths. Age specific mortality rates were calculated by dividing the number of deceased people in the three years of follow up by the observed number of person years in that period. The mortality per five year age group was calculated as 1e–. (mortality rate. . Next, we estimated mortality rates as had been present for earlier. /1000) * time. birth cohorts in former time. To this end we assumed the characteristics of a stable population when interpreting the demographic data as observed during our follow up period, i.e. a constant birth- and mortality rate, and a negligible migration. Such a stable population is assumed to have an intrinsic growth rate between 1% and 3% (Siegel and Swanson, 2004). The expected number of deaths per age group was calculated by subtracting the number of living subjects in the successive age group from the number of subjects in the age group of interest, i.e. attributing the difference in numbers to deaths. Part of the different numbers between age groups could also have been caused by the intrinsic growth rate. Therefore, when estimating earlier mortality rates we modeled. intrinsic. growth. rates. of. 1%,. 2%,. and. 3%. Age. groups. were. assumed to be exposed to a constant intrinsic growth rate for ten years. This. 75.

(8) results in a natural growth of 1.0110, 1.0210, or 1.0310, respectively. Finally, we calculated the life expectancy at birth based on the actual mortality that we observed during our survey. For the calculation of the life expectancy at birth we used the “Life expectancy calculator with 95% confidence intervals” provided by the Eastern Region Public Health Observatory (ERPHO, 2007).. Results At the time of the first census in 2002, there were 17,595 subjects living in the Garu Tempane district of which 47% were male and 53% were female. The median age of the population was 13 years, while 20% percent of the population was younger than five years. Between 2002 and 2005, some 514 (2.9%) subjects moved out of the area and less then 100 (<1%) subjects moved into the area (net emigration was 67 individuals in 2003, 174 in 2004 and 199 in 2005).. In a random series of 2,460 women aged over 25 years the mean number of deliveries was 6.3 (SE 0.05) and 7.7 (SE 0.07) for women aged 45 years and over. The annual number of newborns lowered from 564 (305 male/259 female) in 2003 to 445 in 2005 (257 male/ 188 female).. Observed mortality During a period of three years follow up 578 subjects had died. 263 in 2003, 159 in 2004 and 156 in 2005.. The subjects under observation contributed to. a total number of 54 731 person years. Based on the observed number of deaths in each 5-year age group an abridged life table was constructed (table 1). Mortality in the age group 0 through 4 years was 6,0 per 1000 person years per decade, and in the age group 5 – 9 years it was 2,0 per 1000 person years. Survival probability in the first decade of life was calculated at 0.94, and 0.98 in the second decade. Cumulative survival probability up to age 20 thus amounts to 92%. This level of mortality at childhood and adolescence indicates a transitional status.. Expected mortality We calculated mortality rates also under the assumption that the population would. have. distribution. been as. in. a. stable,. observed. pre-transitional. during. our. study. status,. period.. based. on. the. Calculations. age. were. performed assuming intrinsic growth rates of 1%, 2%, and 3% (table 2). To avoid effects of heaping we used a ten-year abridged life table.. The expected mortality under stable conditions in the age group 0 through 9 years assuming an intrinsic growth rate of 1 % was 404 per 1000 person years and with an 3%. intrinsic growth rate. it was estimated 281 per 1000. 76. person.

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16) Table 1. Observed mortality, survival, and cumulative survival in the Garu-Tempane District 2002 – 2005. .

(17)

(18)

(19) ! " $ %## " ## $ $ $ % "# $ "" ! "" !!# ""$ $% " $! # $ $ # " $ % ## "! "" ##$ ! !! $ !% " "" % %$ ! .

(20)

(21) # ! "% "" ! $ %! # "# !#. $ $% $% $% $! $% $# $! $# $" $ $ %# %# % #$ %. $ $ $ %$ %# % % !% ! ! # $ " .

(22)

(23)

(24)

(25)

(26)

(27) Table 2. Mortality in the Garu-Tempane district assuming the population in a stable, pre-transitional status. .

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35) !

(36)

(37) . !

(38) " !

(39) ". !

(40) #" ( )* ( + +, ,, +,- . ( +-.- ,-) +, *, ( ( -) ). .. ++( )+ ,- *, .(. ,,( (. +( (. ) - --( .)( - .*. . ))( * . )+ ,) -.. **( --- ++) -,, -, ,-( ./ ( (. $. $. $.

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

(50)

(51)

(52)

(53)

(54)

(55)

(56)

(57)

(58) !

(59) " #$

(60) %

(61)

(62)

(63) .

(64)

(65)

(66)

(67)

(68)

(69)

(70)

(71) $ %

(72)

(73)

(74)

(75)

(76)

(77) &

(78)

(79)

(80)

(81)

(82) '

(83)

(84) . 77. .

(85) years. In the age group 10 through 19 years, these figures were 340 and 201 respectively. Beyond age 20, mortality could not always be estimated under the assumption of a 3% population growth rate, and less so when growth rate was assumed to be 2%. The age structure of the population thus suggests that in the recent past intrinsic growth were closer to 1 to 2% than to 2 to 3%.. A comparison of the observed and expected mortality over all age categories is graphically displayed in Figure 1. At young age, expected mortality under the assumption of a pre-transitional status was one to two magnitudes higher than actually was observed in the period 2002 through 2005. This holds for an intrinsic population growth rate of both 1% and 2%. At old age, observed. . . . . . . .

(86) . . . . . .

(87) . Figure. .

(88)

(89)

(90)

(91)

(92)

(93)

(94) . 1. Log mortality curves (Gompertz curves) in the Garu-Tempane district.. Expected mortality refers to estimated mortality assuming a stable population in a pre-transitional status.. 78.

(95) and expected mortality were much closer and became similar under the assumption of a population growth rate of 1%. Comparison with growth domestic product We calculated the life expectancy at birth for the Garu Tempane area, based on the observed mortality in the period 2002-2005. The life expectancy was calculated at 59.4 years (95% CI 57.3 – 61.5). We compared this with the life expectancies and growth domestic product (GDP) of all countries of the world (UNDP, 2003). The life expectancy appears exceptionally high relative to the GDP (figure 2).. . . !"#!$. . . . . . . . . . . .

(96)

(97) . Figure. 2. Life expectancy at birth for all countries in the world dependent on the. Gross Domestic Product (GDP) in US Dollar. The round dot indicates the GaruTempane district.. 79.

(98) Discussion Here we have presented data on mortality patterns in a distinct population within the remote, agricultural area in the Upper East of Ghana. We found far lower than expected mortality at very young age, less so at age 20 through 50 years, and equal to expected mortality from age 50 years onwards. These mortality patterns indicate that this population is in an early transitional status. moving. from. high. mortality. in. childhood. to. a. predominance. of. mortality in late life. Based on actual mortality data, life expectancy in the area is at present considerably higher than in other regions of Ghana (Ghana Statistical. Service,. 2004).. Moreover,. life. expectancy. in. the. area. is. also. considerably higher when compared to similar areas or countries, like the Nouna district in Burkina Faso which has a life expectancy at birth of 47 years (Yazoumé, 2000).. For studying mortality patterns in the area, we had to estimate calendar age for all the subjects within the area as any form of municipal registries are missing. Estimating age carries the risk of misclassification by heaping and misreporting (Spiegelman, 1968). Heaping is the clustering of estimated age with a digit preference for ages ending at 0 and 5. We indeed have observed this type of age preference in our study. But, the observed mortality is not affected by this since we have a follow up period of less than five years. However, for the calculation of the expected mortality the categorization into 5 year age groups was problematic since more subjects were classified in the age groups ending at 0 compared to the age groups ending at 5. Therefore we used 10 year age groups for the expected mortality. Next to heaping there are different. types. of. misreporting,. random. misreporting. and. systematic. misreporting. Random misreporting of age will obscure the differences in mortality between age groups but the data presented here suggest that this will not play a major role. Systematic under- or overestimating age, called shifting, can have distorted the results, especially when comparing the data to external standards. We have not made use of external standards. However, in this population under adverse conditions it may be expected that we have overestimated the age of especially the older people. This will have resulted in a lower mortality in the higher age categories but this will not have affected. our. conclusions. because. we. have. only. compared. observed. and. expected mortality within the same age categories. Among other possible age groups that are susceptible for the effect of shifting are women around the age of giving first child birth. Women who had given birth under twenty years of age might have been estimated over twenty, because they had given birth. We have contentiously tried to reduce this tendency while gathering the information during our field work.. 80.

(99) Over a period of three years we have annually registered the newborns and those who had died during the last year. Different from the subjects who had deceased, it is unlikely that all newborns have been reported. Especially the newborns who were born and died between two of our visits may have been missed.. This might result in an underestimation of neonatal and infant. mortality. that. we. have. observed. in. the. population.. Diallo. et. al.. (1996). showed in a study with similar methods in Burkina Faso, that underreporting of. neonatal. and. infant. deaths. was. limited. to. 2. percent.. Although. underreporting favors the interpretation of the data, the difference between the. observed. en. expected. mortality. is. too. large. to. dispute. that. an. epidemiologic transition is going on. When analyzing demographic data of developing countries, it is common to estimate child mortality from the number of children ever born and children surviving (UNDP, 1983). We were unable to use this method as an internal check of the quality of our data as we do not have this type of data for women below 25 years of age.. There are two explanations why the mortality that we have observed is so much lower than the expected mortality when the population would still be in a pre-transitional state, that are an increased birth rate or migration. It is unlikely that the discrepancy can be explained by an increase of the number of newborns. Women in the region still experience maximum fecundity and our field interviews have indicated there is no birth control. During our annual. follow. newborns. up. whereas. visits the. we. have. number. of. observed. a. women. of. decrease child. in. the. bearing. number. age. had. of not. changed. This however, is a very recent phenomenon, as the women at reproductive age show very high maternal fertility rates that are compatible with a pre transitional status. Second, it is unlikely that an increased ruralurban migration has had a major influence on the mortality figures since migration from the Garu-Tempane district appears to be very low (less than 1% per year). Most of the migrants leave the area to find work and therefore expected. mortality. in. the. age. group. 10-19. years. may. be. slightly. overestimated. Taken together, the lower observed mortality compared to the expected mortality is likely to be a true finding and indicates the start of an epidemiologic transition in the area. The recent decrease in the number of newborns is in line with this observation. Based upon our findings we come to the conclusion that the results of the Demographic Health Survey 2003 are likely to be true. The Upper East Region is experiencing an overall low mortality and remarkable low child mortality now. These are strong indications that the area is in epidemiologic transition.. Until now there is no clear explanation for the observed epidemiological transition. Basically there are three major factors that may contribute to a. 81.

(100) rapid transition: nutrition, medical care (including vaccination) and hygiene and sanitation. The nutrition status of the population is still poor, although the introduction of fertilizers has definitely improved the nutrition status of some families in the area, it does not explain the total decline in mortality. Overall, the food situation has not improved considerably over the last decades and it does not explain why child mortality in the relatively poor and dry Upper East region is much lower than in other, more developed and more fertile parts. of. Ghana.. We. conclude. that. improvement. of. the. nutrition. status-. although limited – may play a part in the transition, but does not fully explain the strongly reduced mortality.. Secondly, the fact that the healthcare system is still in its infancy and has remained virtually unchanged over the last two decades, could explain why the overall mortality of the population has not yet improved, but does not explain why the mortality in the youngest age categories did. We hypothesize that the introduction of mass vaccination programs might be an important driver. of. the. decrease. in. mortality. at. young. age.. The. first. structured. vaccination program in the region (the Expand Program on Immunization (EPI)) started around 1985 and was aimed at vaccinating children under the age of 12 years. As shown, the effects on mortality are indeed strongest in the first two decades of life. In our remote research area bed-nets are almost absent and those available are not very well kept and the effect of its use on the total decline in mortality is negligible.. Finally, we consider that the introduction of bore holes around 1975 might have had some impact. Since sewage disposal systems, water mains, and school hygiene are not introduced yet, it is not very likely that hygiene improvements are the main reason for the reduced mortality.. The strength of our study is our ability to consequently follow up, and therefore having the ability to calculate mortality in a remote and under developed area where municipal registrations are absent. By comparing the observed and expected mortality, we were able to show that the area has started. an. epidemiologic. transition. confirming. the. data. from. the. 2003. Demographic Health Survey in this area.. Although the mechanism behind the reduced mortality is not completely revealed in this study, the results support the idea that, in an area with the very. low. per. capita. income. of. less. than. 100. dollar. per. year. interventions are sufficient to dawn an epidemiologic transition.. 82. low. cost.

(101) Conflict of Interest: None. Ethical Clearance: The Medical Ethical Committee of the Ghana Health Service, as well as the Medical Ethical Committee of the Leiden University Medical Center approved the studies. Witnessed observed informed consent was obtained from all participants.. Acknowledgement: We thank Juventus Ziem, University of Development Studies,. Tamale,. Ghana,. for. his. kind. co-operation. and. provision. of. the. original demographic database.. Authors’ contributions : JJM, JA and RGJW designed the study protocol, JJM, JA and DP carried out the field work in Ghana, JJM, AJMdC, DP and RGJW analyzed the results and wrote the article. All authors have read and approved the final manuscript. JJM and RGJW are the guarantors of the paper.. 83.

(102) References. Adjuik, M., Smith, T., Clark, S., Todd, J., Garrib, A., Kinfu, Y., Kahn, K., Mola, M., Ashraf, A., Masanja, H., Adazu, K., Sacarlal, J., Alam, N., Marra, A., Gbangou, A., Mwageni, E., Binka, F., 2006. Cause-specific mortality rates in sub-saharan africa and bangladesh. Bull. World Health Organ. 84, 181-188. Diallo, D.A., Habluetzel, A, Esposito, F., Cousens, S.N., 1996. Comparison of two methods for assessing child mortality in areas without comprehensive registration systems. Trans. R. Soc. Trop. Med. Hyg. 90, 610-613. Breiman, R.F., Streatfield, P.K., Phelan, M., Shifa, N., Rashid, M., Yunus, M., 2004.. Effect. of. infant. immunization. on. childhood. mortality. in. rural. bangladesh: analysis of health and demographic surveillance data. Lancet 364, 2204-2211. Expanded Program on Immunization (EPI), 2005. Annual report 2005 upper east region. Ghana Health Service UER Office Bolgatanga, Ghana. ERPHO, 2004 Garmin.. http://www.erpho.org.uk. (accessed September 2007).. Garmin. 5. Mapsource. Global. Positioning. System. 2001-2002;. Olathe, Kansas KS66062-3426 USA. Ghana. Statistical. Service,. 2004.. Demographic. and. health. survey. 2003.. Noguchi Memorial Institute for Medical Research, Legon, Ghana / Cleverton Maryland, USA. Ghana VAST Study Team. 1993. Vitamin A supplementation in northern Ghana:. effects. on. clinic. attendances,. hospital. admissions,. and. child. mortality. Lancet 342, 7-12. Hammer, G., Somé, F., Müller, O., Kynast-Wolf, G., Kouyaté, B., Becher, H., 2006. Pattern of cause-specific childhood mortality in a malaria endemic area of Burkina Faso. Malar. J. 5, 47. Huynen, M.M., Vollebregt, L., Martens, P., Benavides, B.M., 2005. The epidemiologic transition in Peru. Rev. Panam. Salud. Publica. 17, 51-59. International Fund for Agricultural Development, 2006. Upper east region land conservation and smallholder rehabilitation project (LACOSREP) – Phase II interim evaluation. Report no. 1757-GH17. Van Landingham, M., Hirschman, C., 2001. Population pressure and fertility in pre-transition thailand. Popul. Stud.55, 233-248. Moniruzzaman, S., Andersson, R., 2005. Relationship between economic. 84.

(103) development and risk of injuries in older adults and the elderly: A global analysis. of. unintentional. injury. mortality. in. an. epidemiologic. transition. perspective. Eur. J. Public Health.15,.454-458. Moore, D., Castillo, E., Richardson, C., Reid, R.J., 2003. Determinants of health. status. and. the. influence. of. primary. health. care. services. in. Latin. America, 1990-98. Int. J. Health Plann. Manage. 18,.279-292. Omran, A., 1971. The epidemiologic transition. A theory of the epidemiology of population change. Milbank Mem. Fund. Q. 49,.509-538. Siegel, J., Swanson, D., 2004. The methods and materials of demography. San Diego, California: Elsevier Academic Press. Spiegelman,. M.,. 1968.. Introduction. to. demography.. Cambridge,. MA. :. Harvard University Press. Storey, P.A., Spannbrucker, N., Agongo, E.A., van Lieshout, L., Ziem, J.P., 2002. Intraobserver and interobserver variation of ultrasound diagnosis of oesophagostomum bifurcum colon lesions. Am. J. Trop. Med. Hyg..67,.680683. UNDP, 2003. United. http://hdr.undp.org/reports [accessed may 10, 2007].. Nations,. estimation.. 1983.. Manual. Department. of. X.. Indirect. International. techniques. Economic. for. and. demographic. Social. Affairs. Population Division. New York: United Nations Publication. Vaugelade, J., Pinchinat, S., Guiella, G., Elguero, E., Simondon, F., 2004. Non-specific effects of vaccination on child survival: prospective cohort study in Burkina Faso. BMJ. 329,1309-1311. WHO, 2003. WHO statistics. United nations world population prospects: the 2002 revision. New York: UN, Department of Economic and Social Affairs, Population Division. Wolleswinkel-van. den. Bosch,. J.H.,. Looman,. C.W.,. Van. Poppel,. F.W.,. Mackenbach, J.P., 1997. Cause-specific mortality trends in the Netherlands, 1875. –. 1992:. a. formal. analysis. of. the. epidemiologic. transition.. Int.. J.. Epidemiol. 26,.772-781. Yazoumé Y., Sanou,A., Gbangou, A., Kouyaté, B., 2000. Nouna demographic surveillance system. .Nouna Health Research Centre. Burkina Faso. Ziem, J.B., Olsen, A., Magnussen, P., Horton, J., Spannbrucker, N., Yrlifari, L., Nana Biritwum, K., Polderman, A.M., 2005. Annual mass treatment with albendazole might eliminate human oesophagostomiasis from the endemic focus in northern Ghana. Trop. Med. Int. Health. 11,.1759-1763.. 85.

(104)

No results found