Prepare to become a physician, build your knowledge, lead a healthcare organization, and advance your career with NEJM Group’s information and services.
It has been speculated that in high transmission settings, malaria control in early childhood (<5 years) may delay the acquisition of functional immunity and shift child mortality from younger to older.
We used data from a 22-year prospective cohort study in rural southern Tanzania to estimate the association between early use of treated nets and survival to adulthood.All children born in the study area between 1 January 1998 and 30 August 2000 were invited to participate in the longitudinal study from 1998 to 2003.Adult survival outcomes were validated in 2019 by community outreach and mobile phone calls.We used Cox proportional hazards models to estimate the association between early childhood use of treated nets and survival in adulthood, adjusted for potential confounders.
A total of 6706 children were enrolled.In 2019, we verified vital status information for 5983 participants (89%).According to reports from early community outreach visits, about a quarter of children never slept under a treated net, half slept under a treated net at some point, and the remaining quarter always slept under a treated net. Sleep under treated mosquito nets.The reported hazard ratio for death was 0.57 (95% confidence interval [CI], 0.45 to 0.72).less than half the visits.The corresponding hazard ratio between age 5 and adulthood was 0.93 (95% CI, 0.58 to 1.49).
In this long-term study of early malaria control in high-transmission settings, the survival benefits of early use of treated nets persisted into adulthood.(Funded by the Eckenstein-Geigy Professorship and others.)
Malaria remains the leading cause of disease and death globally.1 Of the 409,000 malaria deaths in 2019, more than 90% occurred in sub-Saharan Africa, and two-thirds of the deaths occurred in children under the age of five.1 Insecticide-treated nets have been the backbone of malaria control since the 2000 Abuja Declaration 2 .A series of cluster-randomized trials conducted in the 1990s showed that treated nets had a substantial survival benefit for children under 5 years of age.3 Mainly due to large-scale distribution, 2019.1 46% of malaria-risk populations in sub-Saharan Africa sleep in treated mosquito nets
As evidence emerged in the 1990s of the survival benefit of treated nets for young children, it is hypothesized that the long-term effects of treated nets on survival in high-transmission settings will be lower than the short-term effects, and may even be negative, due to the net gain of acquiring functional immunity. related delays.4-9 However, published evidence on this issue is limited to three studies from Burkina Faso, Ghana,11 with follow-up of no more than 7.5 years and Kenya.12 None of these publications showed evidence of a shift in child mortality from young to old age as a result of early childhood malaria control.Here, we report data from a 22-year prospective cohort study in rural southern Tanzania to estimate the association between early childhood use of treated mosquito nets and survival in adulthood.
In this prospective cohort study, we followed children from early infancy through adulthood.The study was approved by the relevant ethical review boards in Tanzania, Switzerland and the United Kingdom.Parents or guardians of young children gave verbal consent to data collected between 1998 and 2003.In 2019, we obtained written consent from participants interviewed in person and verbal consent from participants interviewed by telephone.The first and last authors vouch for the completeness and accuracy of the data.
This study was conducted at the Ifakara Rural Health and Demographic Surveillance Site (HDSS) in the Kilombero and Ulanga regions of Tanzania.13 The study area initially consisted of 18 villages, which were later divided into 25 (Fig. S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org).All children born to HDSS residents between January 1, 1998, and August 30, 2000 participated in the longitudinal cohort study during home visits every 4 months between May 1998 and April 2003.From 1998 to 2003, participants received HDSS visits every 4 months (Fig. S2).From 2004 to 2015, the survival status of participants known to live in the area was recorded in routine HDSS visits.In 2019, we conducted follow-up surveys through community outreach and cell phones, verifying the survival status of all participants, independent of place of residence and HDSS records.The survey relies on family information provided at enrollment.We created a search list for each HDSS village, showing the first and last names of all former family members of each participant, along with the date of birth and the community leader responsible for the family at the time of registration.In meetings with local community leaders, the list was reviewed and other community members were identified to help track.
With the support of the Swiss Agency for Development and Cooperation and the Government of the United Republic of Tanzania, a programme to conduct research on treated mosquito nets was established in the study area in 1995.14 In 1997, a social marketing programme aimed at distributing, promoting and recovering part of the cost of nets, introduced net treatment.15 A nested case-control study showed that treated nets were associated with a 27% increase in survival in children aged 1 month to 4 years (95% confidence interval [CI], 3 to 45). 15
The primary outcome was survival verified during home visits.For participants who have died, age and year of death were obtained from parents or other family members.The main exposure variable was the use of mosquito nets between birth and 5 years of age (“net use in early years”).We analyzed network availability at the individual usage and community levels.For personal use of mosquito nets, during each home visit between 1998 and 2003, the child’s mother or caregiver was asked if the child’s mother or caregiver had slept under the net the previous night, and if so, if and when the net was insecticide- Handling or washing.We summarized each child’s early-year exposure to treated nets as the percentage of visits in which children were reported to be sleeping under treated nets.For village-level treatment network ownership, we combined all household records collected from 1998 to 2003 to calculate the proportion of households in each village that owned at least one treatment network by year.
Data on malaria parasitemia were collected in 2000 as part of a comprehensive surveillance program for antimalarial combination therapy.On 16 May, in a representative sample of HDSS families, parasitemia was measured by thick film microscopy in all family members 6 months or older through July 2000, 2001, 2002, 2004, 2005 Year and 2006.16
To maximize data quality and completeness of follow-up in 2019, we recruited and trained a team of experienced interviewers who already had extensive local knowledge.For some families, information on caregiver education, family income, and time to medical facility was not available.Multiple imputation using chain equations was used to account for missing covariate data in our primary outcome.All variables listed in Table 1 were used as predictors for these imputations.An additional full case study was performed to ensure that the results were not sensitive to the imputation method chosen.
Initial descriptive statistics included mean follow-up visits and mortality by sex, year of birth, caregiver education, and household income category.Mortality is estimated as deaths per 1000 person-years.
We provide data on how network coverage has changed over time.To illustrate the relationship between village-level household ownership of treated bed nets and local malaria transmission, we created a scatterplot of village-level treated bed net coverage and village-level parasitic disease prevalence in 2000.
To estimate the association between net use and long-term survival, we first estimated unadjusted standard Kaplan-Meier survival curves comparing children who reported sleeping under the treated net during at least 50% of early visits with those survival outcome.Children reportedly slept under treated mosquito nets in less than 50% of early visits.The 50% cutoff was chosen to match the simple “most of the time” definition.To ensure that the results were not affected by this arbitrary truncation, we also estimated unadjusted standard Kaplan-Meier survival curves comparing children who always reported sleeping under the treated net with those who never reported sleeping under the treated net Survival outcomes of children under the net. We estimated unadjusted Kaplan-Meier curves for these contrasts after the entire period (0 to 20 years) and early childhood (5 to 20 years).All survival analyses were limited to the time between the first survey interview and the last survey interview, which resulted in left truncation and right censoring.
We used Cox proportional hazards models to estimate three main contrasts of interest, conditional on observable confounders—first, the association between survival and the percentage of visits in which children reportedly slept under treated nets; second, Differences in survival between children who used treated nets at more than half of their visits and those who used treated nets at less than half of their visits; third, differences in survival between children always reported sleeping at their early visits Under treated mosquito nets, the children never reported sleeping under treated nets during these visits.For the first association, the visit percentage is analyzed as a linear term.A martingale residual analysis was performed to confirm the adequacy of this linearity assumption.Schoenfeld residual analysis17 was used to test the proportional hazards assumption.To account for confounding, all multivariate estimates for the first three comparisons were adjusted for household income category, time to nearest medical facility, caregiver’s education category, child’s sex, and child’s age.born.All multivariate models also included 25 village-specific intercepts, which allowed us to exclude systematic differences in unobserved village-level factors as potential confounders.To ensure the robustness of the presented results with respect to the chosen empirical model, we also estimated two binary contrasts using kernels, calipers and exact matching algorithms.
Given that early use of treated nets could be explained by unobserved household or caregiver characteristics such as health knowledge or an individual’s ability to access medical services, we also estimated a village-level model as a fourth contrast.For this comparison, we used village-level average household ownership of treated nets (input as a linear term) in the first 3 years in which children were observed as our primary exposure variable.Village-level exposure has the advantage of being less dependent on individual or household-level covariates and should therefore be less affected by confounding.Conceptually, increasing village-level coverage should have a greater protective effect than increasing individual coverage due to greater effects on mosquito populations and malaria transmission.18
To account for village-level net treatment as well as village-level correlations more generally, standard errors were calculated using Huber’s cluster-robust variance estimator.Results are reported as point estimates with 95% confidence intervals.The widths of the confidence intervals are not adjusted for multiplicity, so the intervals should not be used to infer established associations.Our primary analysis was not prespecified; therefore, no P-values were reported.Statistical analysis was performed using Stata SE software (StataCorp) version 16.0.19
From May 1998 to April 2003, a total of 6706 participants born between January 1, 1998 and August 30, 2000 were included in the cohort (Figure 1).Enrollment ages ranged from 3 to 47 months, with a mean of 12 months.Between May 1998 and April 2003, 424 participants died.In 2019, we verified the vital status of 5,983 participants (89% of the enrollment).A total of 180 participants died between May 2003 and December 2019, resulting in an overall crude death rate of 6.3 deaths per 1000 person-years.
As shown in Table 1, the sample was gender-balanced; on average, children were enrolled just before turning one year old and followed for 16 years.Most caregivers have completed primary education, and most households have access to tap or well water.Table S1 provides more information on the representativeness of the study sample.The observed number of deaths per 1000 person-years was lowest among children with highly educated caregivers (4.4 per 1000 person-years) and highest among children who were more than 3 hours away from a medical facility (9.2 per 1000 person-years) and Among households lacking information on education (8.4 per 1,000 person-years) or income (19.5 per 1,000 person-years).
Table 2 summarizes the main exposure variables.About a quarter of the study participants reportedly never slept under a treated net, another quarter reported sleeping under a treated net at each early visit, and the remaining half slept under some but not all Reported sleeping under treated mosquito nets at the time of visit.The proportion of children who always slept under treated mosquito nets increased from 21% of children born in 1998 to 31% of children born in 2000.
Table S2 provides more details on overall trends in network usage from 1998 to 2003.Although it was reported that 34% of children slept under treated mosquito nets the night before in 1998, by 2003 that number had increased to 77%.Figure S3 shows the net frequency of use treated early in life.Figure S4 shows the high variability of ownership, with less than 25% of households having treated nets in Iragua village in 1998, while in Igota, Kivukoni and Lupiro villages, more than 50% of households had treated nets in the same year.
Unadjusted Kaplan-Meier survival curves are shown.Panels A and C compare the (unadjusted) survival trajectories of children who reported using treated nets for at least half the number of visits to those who used less frequently.Panels B and D compare children who never reported sleeping under treated nets (23% of the sample) with those who always reported sleeping under treated nets (25% of the sample). adjusted) track.The inset shows the same data on an enlarged y-axis.
Figure 2 Comparison of participants’ survival trajectories to adulthood based on early use of treated nets, including survival estimates for the entire period (Figures 2A and 2B) and survival curves conditioned on survival to 5 years of age (Figures 2C and 2D).A total of 604 deaths were recorded during the study period; 485 (80%) occurred in the first 5 years of life.Mortality risk peaked in the first year of life, declined rapidly until age 5, then remained relatively low, but increased slightly at about age 15 (Fig. S6).Ninety-one percent of participants who consistently used treated nets survived to adulthood; this was also the case for only 80% of children who did not use treated nets early on (Table 2 and Figure 2B).Parasite prevalence in 2000 was strongly negatively correlated with treated bed nets owned by households of children under 5 years (correlation coefficient, ~0.63) and children 5 years of age or older (correlation coefficient, ~0.51) (Fig. S5). ).
Each 10-percentage-point increase in early use of treated nets was associated with a 10% lower risk of death (hazard ratio, 0.90; 95% CI, 0.86 to 0.93), provided the full set of caregivers and household covariates were as well as the village fixed effects (Table 3 ).Children who used treated nets at earlier visits had a 43% lower risk of death compared with children who used treated nets at less than half of their visits (hazard ratio, 0.57; 95% CI, 0.45 to 0.72).Likewise, children who always slept under treated nets had a 46% lower risk of death than children who never slept under nets (hazard ratio, 0.54; 95% CI, 0.39 to 0.74).At the village level, a 10-percentage-point increase in treated bed net ownership was associated with a 9% lower risk of death (hazard ratio, 0.91; 95% CI, 0.82 to 1.01).
The use of treated nets during at least half of the early-life visits was reported to be associated with a hazard ratio of 0.93 (95% CI, 0.58 to 1.49) for death from age 5 to adulthood (Table 3).In the initial period from 1998 to 2003, when we adjusted for age, caregiver education, household income and wealth, year of birth and village of birth (Table S3).
Table S4 shows surrogate propensity scores and exact match estimates for our two binary exposure variables, and the results are nearly identical to those in Table 3.Table S5 shows differences in survival stratified by number of early visits.Despite relatively few observations for at least four early visits, the estimated protective effect appears to be greater in children with more visits than in children with fewer visits.Table S6 shows the results of the full case analysis; these results are nearly identical to those of our main analysis, with slightly higher precision for the village-level estimates.
Although there is strong evidence that treated nets can improve survival in children under 5 years of age, studies of long-term effects remain scarce, especially in areas with high transmission rates.20 Our results suggest that children have significant long-term benefits from using treated nets.These results are robust across broad empirical norms and suggest that concerns about increased mortality in later childhood or adolescence, which could theoretically be due to delayed functional immune development, are unfounded.Although our study did not directly measure immune function, it can be argued that survival into adulthood in malaria-endemic areas is itself a reflection of functional immunity.
Strengths of our study include the sample size, which included more than 6500 children; the follow-up time, which was a mean of 16 years; the unexpectedly low rate of loss to follow-up (11%); and the consistency of results across analyses.The high follow-up rate may be due to an unusual combination of factors, such as the widespread use of mobile phones, the cohesion of the rural community in the study area, and the deep and positive social ties developed between researchers and local people.Community via HDSS.
There are certain limitations of our study, including the lack of individual follow-up from 2003 to 2019; no information on children who died before the first study visit, which means that cohort survival rates are not fully representative of all births in the same period; and observational analysis.Even if our model contains a large number of covariates, residual confounding cannot be ruled out.Given these limitations, we suggest that further research is needed on the impact of long-term continued use of bed nets and the public health importance of untreated bed nets, especially given current concerns about insecticide resistance.
This long-term survival study related to early childhood malaria control shows that with moderate community coverage, the survival benefits of insecticide-treated bed nets are substantial and persist into adulthood.
Data collection during the 2019 follow-up by Prof. Eckenstein-Geigy and support from 1997 to 2003 by the Swiss Agency for Development and Cooperation and the Swiss National Science Foundation.
The disclosure form provided by the authors is available with the full text of this article at NEJM.org.
The data sharing statement provided by the authors is available with the full text of this article at NEJM.org.
From Swiss Tropical and Public Health Institute and University of Basel, Basel, Switzerland (GF, CL); Ifakara Health Institute, Dar es Salaam, Tanzania (SM, SA, RK, HM, FO); Columbia University, New York Mailman School of Public Health (SPK); and London School of Hygiene and Tropical Medicine (JS).
Dr. Fink can be contacted at [email protected] or at the Swiss Institute for Tropical and Public Health (Kreuzstrasse 2, 4123 Allschwil, Switzerland).
1. World Malaria Report 2020: 20 Years of Global Progress and Challenges.Geneva: World Health Organization, 2020.
2. World Health Organization.The Abuja Declaration and Action Plan: Extracts from the Roll Back Malaria Africa Summit.25 April 2000 (https://apps.who.int/iris/handle/10665/67816).
3. Pryce J, Richardson M, Lengeler C. Insecticide-treated mosquito nets for malaria prevention.Cochrane Database System Rev 2018;11:CD000363-CD000363.
4. Snow RW, Omumbo JA, Lowe B, et al.Association between the incidence of severe malaria in children and the level of Plasmodium falciparum transmission in Africa.Lancet 1997;349:1650-1654.
5. Experiments by Molineaux L. Nature: What are the implications for malaria prevention?Lancet 1997;349:1636-1637.
6. D’Alessandro U. Malaria severity and level of Plasmodium falciparum transmission.Lancet 1997;350:362-362.
8. Snow RW, Marsh K. Clinical Malaria Epidemiology in African Children.Bull Pasteur Institut 1998;96:15-23.
9. Smith TA, Leuenberger R, Lengeler C.Child mortality and malaria transmission intensity in Africa.Trend Parasite 2001;17:145-149.
10. Diallo DA, Cousens SN, Cuzin-Ouattara N, Nebié I, Ilboudo-Sanogo E, Esposito F. Insecticide-treated curtains protect child mortality in West African populations for up to 6 years.Bull World Health Organ 2004;82:85-91.
11. Binka FN, Hodgson A, Adjuik M, Smith T. Mortality in a seven-and-a-half-year follow-up trial of insecticide-treated mosquito nets in Ghana.Trans R Soc Trop Med Hyg 2002;96:597-599.
12. Eisele TP, Lindblade KA, Wannemuehler KA, et al.Effects of continued use of insecticide-treated bed nets on all-cause mortality in children in areas of western Kenya where malaria is highly perennial.Am J Trop Med Hyg 2005;73:149-156.
13. Geubbels E, Amri S, Levira F, Schellenberg J, Masanja H, Nathan R. Introduction to the Health and Population Surveillance System: Ifakara Rural and Urban Health and Population Surveillance System (Ifakara HDSS).Int J Epidemiol 2015;44:848-861.
14. Schellenberg JR, Abdulla S, Minja H, et al.KINET: A social marketing programme for the Tanzania Malaria Control Network assessing child health and long-term survival.Trans R Soc Trop Med Hyg 1999;93:225-231.