Background
Sub-Saharan Africa has witnessed an epidemiological transition in the distribution and intensity of malaria transmission since 2000, with a remarkable decline of malaria burden up to 2015 and a resurgence reported in 2016 [
1,
2]. The declining burden of malaria has often been directly attributed to scale-up of interventions including vector control and changing anti-malarial treatment policies [
1‐
4]. However, malaria is influenced by a complex array of environmental, ecological and climate factors and a desire to attribute the changes to intervention may mask this complex interplay of biotic and abiotic factors [
2]. A detailed analysis of factors affecting malaria transmission requires congruent temporal data over long periods [
2]. However, such long-term data which would support stratification of disease burden and setting up targeted control strategies with most impact is commonly lacking and decisions are often made based on limited updated evidence and high levels of uncertainty. Thus, malaria control strategies and interventions have frequently been implemented without proper stratification of current malaria transmission which would have presented a more balanced picture of malaria transmission intensity and disease burden in specific areas, and provide better evidence for more targeted interventions with higher impacts [
5]. Malaria control in Tanzania is of no exception: a paper examining the distribution of insecticide-treated nets (ITNs) showed clustering of ITNs which generally did not reflect patterns of transmission and in some places even an inverse distribution vs. intensity of transmission [
6].
Longitudinal studies in Kenya [
7,
8], Senegal [
9] and Guinea Bissau [
10] have all demonstrated that the changing epidemiology of infection and disease cannot be easily explained by changing coverage of interventions such as vector control alone. While these studies all demonstrate an overall decline in malaria, in Kenya and Guinea Bissau, resurgent risks were documented after 2012 and this was consistent with resurgent risks during the same period in Zimbabwe [
11] and Mozambique [
12,
13]. Furthermore, recent reports have also showed a resurgence of malaria in many countries particularly in 2016, whereby malaria cases increased by over 6 million compared to 2015 [
1,
2].
Despite scarcity of long-term data on malaria prevalence across Tanzania, the consensus that the burden of malaria has declined significantly since 2000 is most likely valid [
14,
15]. However, the decline has been less dramatic in some of the historically high burden northwestern and southern areas of Tanzania [
14‐
16]. In the village of Nyamisati, in the Rufiji River Delta, Pwani region malaria prevalence was > 70% in 1985 and had declined to 5% in 2010 [
17]. In Korogwe district, Tanga region, the prevalence
of Plasmodium falciparum declined from 78% in 2003 to 13% in 2008 in the lowland village and from 25 to 3% in the highland village [
18]. Finally, a study conducted in the two villages of Magoda and Mpapayu in Muheza district, also of Tanga region showed high malaria prevalence (> 68%) from 1992 to 1999 which was followed by a moderate decline from 2000 to 2004 and then, a significant decline from 2008 to 2012 [
19]. The present study is an extension of the latter study where the aims were to describe the most recent trends of
P.
falciparum prevalence in the two villages and examine plausible factors that may have contributed to the changing transmission intensity of
P.
falciparum between 1992 and 2017.
Discussion
Previously, in the villages of Magoda and Mpapayu, Tanga region, Tanzania, a significant decline in the prevalence of
P. falciparum infections was observed between 2000 and 2012 [
19]. Parasite prevalence as low as < 10% was recorded in 2011 and 2012, and there were no individuals among under-fives from Mpapayu who had malaria parasites during the cross sectional survey performed in 2012. The villages had many interventions deployed during the period of 1998–2004 where the prevalence of
P. falciparum declined, which would easily be ascribed to the interventions.
After the end of major research projects in the two villages in 2004, ITNs and LLINs were distributed to the study communities through different national initiatives including nets for pregnant women between 2006 and 2014, and mass distribution in 2011 and 2016 (NMCP, unpublished data). Moreover, the mobile clinic operating in the villages since the mid-1990s was in 2004 replaced with a dispensary, which offered medical services to all members of the community. Thus, the observed decline in malaria prevalence may be attributed to high coverage of ITNs/LLINs and improved case management at the dispensary, including deployment of effective anti-malarials (artemisinin-based combination therapy from 2007), as also described in the previous study [
19]. It is, therefore, tempting to assume that the decline in malaria burden is closely related to the interventions as has been suggested for Zanzibar as well [
35].
A large epidemiological malaria study conducted in Kenya suggests that the main factor responsible for the decline of malaria was the considerable use of SP in the management of uncomplicated malaria, when it replaced chloroquine as standard treatment in the early 2000’s [
8]. Whether SP has caused a similar impact in the study area remains doubtful, as the decline began 1 year before the introduction of SP (in 2001) and high levels of SP resistance was already reported in these villages before SP was introduced, possibly related to previous mass drug administration of the antifolate drug combination, dapsone–pyrimethamine (Maloprim) in Magoda village in 1993–1994 [
22,
36].
Alternatively, an entomological study performed in two villages of the same Tanga region located about 20 km from the present study observed a drastic decline in abundance of
An. gambiae and
An. funestus; from a sampling of 5382
Anopheles sp. in 2004 to merely a total of 14 mosquitoes in 2009 [
37]. The cause of the almost complete disappearance of these mosquitoes is largely unknown as no major interventions in that study areas was done (coverage with bed nets did not exceed 27%). The study does suggest however, that part of the decline was associated with changes in patterns of monthly rainfall [
37]. Whether the decline in
P. falciparum prevalence observed between 2000 and 2012 in the present study sites was affected by lack of malaria vectors during this period is likely, and could possibly be related to the abnormalities in rainfall (amount and patterns), but unfortunately, no entomological surveys were performed during that study period.
From 2013 onwards, there has been a sustained re-emergence of malaria in the study area with an increase in parasite prevalence reaching 25% in 2014 and remaining relatively higher up to 2016 and a remarkable decline in 2017. Re-emergence of malaria in the study area occurred during a period where ITNs/LLINs are expected to work and a lack of protective effect of ITNs/LLINs somewhat reflects similar observations from a Kenyan study, where an increase in malaria prevalence was reported despite recent mass ITNs/LLINs distribution [
8]. Although use of ITNs/LLINs among under-fives in Tanzania declined between 2012 and 2015 (from 72.7 to 54.5%), the level reported during this period was still high enough compared to the relatively low ITNs/LLINs coverage in 2004 (19.5%) and 2008 (32.7%) but with the highest decline in malaria burden (NMCP, unpublished data). Another potential contributing factor to resurgence of malaria might be development of resistance to insecticides used to impregnate the bed nets (even with LLINs), which has been reported in different parts of Tanzania including Muheza district [
38‐
40].
The declining prevalence up to 2012 occurred in the period with overall negative rainfall anomaly suggesting that the decline in parasite prevalence and malaria burden reported in this area (and other parts of the country) was also possibly affected by the variability in the amount and pattern of rainfall directly affecting the Anopheles mosquitoes. These changes in prevalence occurred simultaneously with the interventions (high bed net coverage from 1998 in Magoda and 2001 in Mpapayu, and anti-malarial drugs) applied during the study period, indicating that multiple factors could be responsible for the observed trends. Conversely, the lack of changes in 2016 and a decline in prevalence in 2017 could possibly be attributed to the negative anomaly in 2016 and high bed net coverage (> 92%).
Thus, the decline in the prevalence of
P. falciparum infections (and current resurgence) is most likely due to multiple factors including anti-malarial control interventions, but the exact contribution of the interventions is uncertain and possibly varies through time [
2]. Additionally, factors such as climate, physical environment and socio-economic development are crucial and should be assessed in order to fully comprehend the changes in malaria epidemiology occurring in recent years. Further integrated surveillance is required to provide additional details which could be responsible for the decline as well as resurging transmission of malaria in North-eastern Tanzania as well as in other parts of Tanzania and sub-Saharan Africa [
1], where malaria transmission patterns, socio-economic conditions; and climate are changing fast.
Authors’ contributions
DSI, MML, MNM, MA and ICB conceived of and designed the study; DSI, CIM, BPM, MGC, FF and MML conducted the fieldwork and supervised the laboratory analyses. WT, HM and AK collected and managed rainfall data. BPM and DSI participated in data management and analysis and DSI prepared the initial draft of the manuscript. DSI, BPM, RWS, MA and ICB reviewed and finalized the manuscript. All authors read and approved the final manuscript.
Acknowledgements
Authors wish to thank the survey teams (Method Segeja, Samuel Gesase, Julius Massaga, Daniel Minja, Hamis Msangeni, Julius Mhina, Juma Akida, Samuel Sembuche, Fredy Saguti, Vito Baraka, Sudi Lwitiho, Deo Maiga, Daniel Challe, Athanas Mhina, Paul Martine, Rashidi Madebe, Ezekiel Malecela, Johari Sadi, Zacharia Savael, Juma Tupa, Malimi Chille, Ibrahimu Materego, Francis Chambo, Magreth Munisi, Zaina Maumba, Beatrice Semng’indo, Salome Simba, Rehema Mtibusa, Mwanaidi Mtui, Clementina Kivumbi, Neema Barua, Benson Swai, Silas Msangi, Sylivia Masawe, Tilaus Gustav, Lydia Masawe, Thompson Mwampamba, Hatibu Athumani, Halima Mpambile, Stella Mkandawile, Allen Mrango, Frank Mnango, Seth Nguhu, Zahabu Kauzeni, Simba Athumani, Sudi Said and Salimu Tembo) and the data management team at Tanga (Chris Msokame, Mary Lukindo, Neema Barua, Benson Swai, Gineson Nkya and Fides Mumburi). Study participants, community members, field workers in (Magoda and Mpapayu) and health authorities in Muheza district are acknowledged for the cooperation and support provided to the study teams. Thanks to the Tanzanian Ministry of Health, Community Development, Gender, Elderly and Children (MoHCDGEC) through NMCP and the Medical Stores Department for donating anti-malarials and RDTs for some of the surveys. The Director General of the National Institute for Medical Research (NIMR) supported the study and provided the permission to publish this paper.