INTRODUCTION
When the transmission of malaria was first discovered over a century ago to occur mainly through indoor biting by anopheline mosquitoes, the importance of house structure to malaria risk was recognized. The earliest experiments were conducted by Angelo Celli in the Italian campagna during the 1890s, and the first review on the subject was published in 1931 by the Malaria Commission of the League of Nations and updated only 85 years later by Tusting and others.1–3 Today, housing-based approaches to malaria control are not routinely deployed, although growing evidence supports their effectiveness.4
The National Malaria Control Program of Zambia aims to eliminate malaria within its borders in this decade, yet transmission remains intractably high in the northern wetlands of Luapula Province despite 12 years of concerted efforts.5 To the south, by contrast, malaria remitted in the wake of drought, aggressive vector control, mass drug administration, case management, and improvements in housing over time.6–11
Housing modifications can safeguard against malaria by reducing contact between mosquito vectors and human hosts. Indoor mosquito abundance, mosquito behavior and life span, and uptake of vector control measures vary with housing.12 Mosquito abundance is affected by construction features such as eaves and other structural gaps which are the primary entry points for mosquitoes.13–15 Resting and feeding behaviors fluctuate with ambient temperature and humidity, which vary with the type of building materials.16,17 Metal roofs appear to discourage mosquito resting and reduce mosquito longevity because of higher temperatures than natural roofs; feeding anophelines rely on evaporative cooling during blood meals and risk overheating when ambient temperatures are too high.16,18 Housing features can ease or restrict bed net use, different wall materials can impact the adsorption and residual efficacy of sprayed insecticide (e.g., different insecticides adhere differently to wood versus plaster), and brighter interiors or lighter walls can facilitate personal protection against mosquitoes by increasing the mosquitoes’ visibility.19–21
House structure and its relationship to malaria risk are interconnected with other socioeconomic indicators such as income, education, occupation, household consumption, asset indices, and other metrics.22,23 Studies that seek to characterize or quantify the associations between socioeconomic indicators and malaria risk are complicated by the tortuosity of the causal pathways and multiply connected variables. For example, the relationship between income and education is bidirectional and relies on their temporal relationship: higher education leads to higher earning potential, and, conversely, greater income affords greater educational opportunities.24 Both education and income modify a person’s likelihood or ability to access preventive resources, which in turn influence malaria risk,25–30 and both are correlated with a person’s occupation and the type of house in which they live.23,31,32 Models that do not account for the mediating relationships across variables can miss or misestimate important associations.33,34
Previous studies of housing and malaria in sub-Saharan Africa focused on community-based populations.3 Here, we investigated housing and other socioeconomic variables to assess their potential as risk factors for malaria within a clinical population. We hypothesized that better housing reduces the odds of malaria independent of other sociodemographic factors, and that house structure plays an important role in the pathways that relate income and education to malaria risk.
METHODS
Study design.
This was a cross-sectional study of child and adult patients presenting with acute febrile illness to rural health centers in Nchelenge District, Luapula Province. Patients were recruited during the rainy season from November to December 2017 for a validation study of a novel isothermal PCR diagnostic test for Plasmodium spp. infection (data not shown).35 Participants provided informed consent. The study protocol was approved by the Ethics Review Committee of the Tropical Diseases Research Centre in Ndola, Zambia.
Study site.
Recruitment was from two rural health centers in Nchelenge District, a high malaria-transmission area that encompasses wetlands along Lake Mweru, bordering the Democratic Republic of Congo.36 The local economy is largely agrarian and fishing based. Local building materials include wooden pole-and-straw construction for walls, sunbaked mud bricks, fired bricks, mixed cement, dried straw thatch, and corrugated metal. Eaves, the ventilation gaps between roof and wall that provide a path of entry for mosquitoes, are present in most houses with thatch roofs but not in those with metal roofs (Figure 1).
Photographs of representative house types showing (A) fired-brick walls and metal roof and (B) sunbaked mud brick walls and thatch roof.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Photographs of representative house types showing (A) fired-brick walls and metal roof and (B) sunbaked mud brick walls and thatch roof.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Photographs of representative house types showing (A) fired-brick walls and metal roof and (B) sunbaked mud brick walls and thatch roof.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Malaria is holoendemic, with an average parasite prevalence of 51% year round peaking in the rainy season (October–April).36 Plasmodium falciparum is the vastly predominant species with rare instances of Plasmodium malariae coinfection or mono-infection. The main vector species is Anopheles funestus sensu stricto (s.s.), most abundant inland from the end of the rainy season and throughout the dry season, and An. gambiae s.s., which peaks lakeside during the rains.37,38 Since 2008, indoor residual spraying (IRS) has been conducted annually, targeted to the lakeside area and timed to the onset of the rainy season, but has had little overall impact on transmission.39 Malaria continues to account for upward of 30–40% of admissions to the children’s ward and 40% of pediatric in-hospital deaths.40
Study participants.
Participants were pediatric and adult patients presenting with acute febrile illness. Inclusion criteria were willingness to participate and provision of informed consent (and assent for children younger than 18 years). Exclusion criteria were severe signs or symptoms of malaria or other illness due to the need of these patients to be transferred from the study site to the nearby hospital.
Study procedures.
All participants underwent testing for P. falciparum infection by thick blood smear microscopy, rapid diagnostic test (RDT, Standard Diagnostics Inc., Suwon, Korea), and subsequent PCR targeting P. falciparum species–specific 18s rRNA using whole blood collected as dried blood spots on Whatman 903 protein saver cards (Sigma-Aldrich, St. Louis, MO) according to published methods.41,42 Blood smears were reviewed independently by two expert microscopists. Parasites were counted against 200 leukocytes and densities estimated assuming 8,000 leukocytes per mm3 of blood.
House structure was classified according to composition of wall (straw-and-pole, cement, or brick) and roof (straw thatch and corrugated metal), as reported by participants. Demographic and other details including IRS within the last 6 months, whether the participant slept under an insecticide-treated bed net (ITN), educational background, occupation of the head of household, and monthly household income were also collected. Housing and demographic data were collected via questionnaires administered to the participant (or participant’s legal guardian) by a trained study team member.
Exposure and outcome.
The main exposure of interest was house type defined according to roof and wall composition (thatch or metal roof, straw-and-pole, or brick walls). Other exposures were participant and household sociodemographic features. The primary outcome was PCR-confirmed diagnosis of malaria.
Statistical analysis.
Data were fitted to unadjusted and adjusted logistic regression models for relational and mediation analyses. Baseline characteristics were compared between groups using Student’s t-test for continuous variables in pairwise comparisons or one-way analysis of variance in multiple group comparisons, or Pearson’s χ2 test for dichotomous variables. Adjusted models incorporated posited confounders that differed across house types (P < 0.10). Mediation of the association between income and malaria by housing, and between education and malaria by income, were assessed using procedures developed by Imai et al. to accommodate binary outcomes.34 These methods apply a counterfactual framework that formally distinguishes main effects from mediation effects. The mediation analyses were predicated on the theories that past educational attainment is positively and causally associated with income in the present, and that present income is positively and causally associated with quality of housing (Figure 2). Statistical analyses were carried out using Stata 16.0 (StataCorp, College Station, TX).
Directed acyclic graph of the hypothesized relationships among education, income, housing, and malaria in febrile patients within a highly malarious area of Zambia. Graph A depicts the partial mediation of the association of income with malaria by housing, and graph B shows the partial mediation of the association of education with malaria by income. Odds ratios (ORs) were computed from logistic regression of malaria PCR test positivity on education (primary vs. secondary or higher), daily income (< 2 USD vs. ≥ 2 USD), and housing (thatch vs. metal roof). Percent contributions were calculated from mediation analyses conducted according to Imai et al.34
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Directed acyclic graph of the hypothesized relationships among education, income, housing, and malaria in febrile patients within a highly malarious area of Zambia. Graph A depicts the partial mediation of the association of income with malaria by housing, and graph B shows the partial mediation of the association of education with malaria by income. Odds ratios (ORs) were computed from logistic regression of malaria PCR test positivity on education (primary vs. secondary or higher), daily income (< 2 USD vs. ≥ 2 USD), and housing (thatch vs. metal roof). Percent contributions were calculated from mediation analyses conducted according to Imai et al.34
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Directed acyclic graph of the hypothesized relationships among education, income, housing, and malaria in febrile patients within a highly malarious area of Zambia. Graph A depicts the partial mediation of the association of income with malaria by housing, and graph B shows the partial mediation of the association of education with malaria by income. Odds ratios (ORs) were computed from logistic regression of malaria PCR test positivity on education (primary vs. secondary or higher), daily income (< 2 USD vs. ≥ 2 USD), and housing (thatch vs. metal roof). Percent contributions were calculated from mediation analyses conducted according to Imai et al.34
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
RESULTS
Participant characteristics.
A total of 282 participants were enrolled in the primary study, of whom 272 provided complete housing data. Table 1 displays participant characteristics according to house type. The median age was 19 years (interquartile range [IQR]: 8–32), and a slight majority (56%) were female. Median monthly household income was Zambian kwacha 250 (IQR: 100–600), equivalent to less than US $1/day. Individuals residing in higher quality houses (metal roof ± solid walls) had on average higher educational attainment and heads of household with higher paying occupations than those in lower quality houses. Almost half (46%) of participants living in the highest quality houses completed secondary school, compared with less than a quarter (22%) in the lowest quality houses (Pearson’s χ2, P < 0.01).
Sociodemographic characteristics of participants according to house type
| Characteristic | Straw-and-pole walls | Cement or brick walls | P-value | ||
|---|---|---|---|---|---|
| Thatch roof | Metal roof | Thatch roof | Metal roof | ||
| n = 154 | n = 7 | n = 24 | n = 87 | ||
| Median age (IQR) (years) | 18 (6–30) | 18 (5–52) | 18 (6–26) | 20 (15–35) | 0.23 |
| Age distribution (years), no. (%) | |||||
| < 5 | 29 (19) | 1 (14) | 3 (13) | 5 (6) | 0.05 |
| 5–15 | 42 (27) | 1 (14) | 8 (33) | 17 (20) | 0.37 |
| 16–29 | 43 (28) | 3 (43) | 8 (33) | 35 (40) | 0.24 |
| 30–39 | 18 (12) | 0 (0) | 1 (4) | 11 (13) | 0.51 |
| ≥ 40 | 22 (14) | 2 (29) | 4 (17) | 19 (22) | 0.41 |
| Female sex, no. (%) | 87 (61) | 2 (33) | 11 (48) | 45 (51) | 0.26 |
| Educational attainment, no. (%) | |||||
| Primary | 75 (49) | 1 (14) | 8 (33) | 31 (36) | 0.06 |
| Secondary | 33 (22) | 4 (57) | 8 (33) | 40 (46) | < 0.01 |
| Tertiary | 0 (0) | 0 (0) | 0 (0) | 4 (5) | 0.04 |
| Occupation, no. (%) | |||||
| Farmer | 82 (54) | 5 (71) | 14 (58) | 35 (40) | 0.09 |
| Fisher | 21 (14) | 1 (14) | 0 (0) | 2 (2) | 0.01 |
| Laborer | 9 (6) | 1 (14) | 0 (0) | 4 (5) | 0.44 |
| Merchant | 13 (9) | 0 (0) | 1 (4) | 6 (7) | 0.74 |
| Office profession | 3 (2) | 0 (0) | 1 (4) | 22 (25) | < 0.01 |
| None | 14 (9) | 0 (0) | 2 (8) | 7 (8) | 0.85 |
| Other | 9 (6) | 0 (0) | 6 (25) | 11 (13) | 0.01 |
| Median income, Zambian kwacha/mo. (IQR) | 200 (100–400) | 500 (200–1,000) | 400 (100–800) | 500 (250–1,700) | < 0.01 |
| Bed net use, no. (%) | 126 (82) | 100 (100) | 23 (96) | 77 (89) | 0.16 |
| Indoor residual spraying in prior 6 months, no. (%) | 95 (62) | 5 (71) | 14 (58) | 38 (44) | 0.04 |
| Mean no. of occupants (SD) | 5.7 (2.2) | 6.6 (1.5) | 5.7 (2.3) | 7.2 (3.1) | < 0.01 |
| Mean no. of rooms (SD) | 3.6 (0.8) | 5.0 (1.3) | 4.0 (1.4) | 5.7 (2.0) | < 0.01 |
IQR = interquartile range. P-values were computed by one-way analysis of variance for continuous variables or Pearson’s χ2 test for dichotomous variables.
Most participants (56%) resided in houses of straw-and-pole walls and thatched roof. The next most common house type (32%) was brick wall and metal roof. There were few living in houses of mixed construction, with 9% in houses with solid walls and thatched roofing, and 3% in houses with natural walls and solid roofing. Prior IRS was more common in lower quality houses (58–71%) than the highest quality houses (44%) (Pearson’s χ2, P = 0.04).
Malaria was diagnosed by PCR in 180 of the 272 participants (66%) (Table 2). Rapid diagnostic test and microscopy detected fewer infections (48% and 38%). Among those with positive blood smears, the median parasite count was 83,200 parasites/μL (range 1,500–2,806,000). One participant had a positive blood smear (17,500 parasites/μL) but negative PCR result, attributed to compromised integrity of the dried blood spot. Eighty participants (29%) had sub-patent malaria defined by a positive PCR result and negative blood smear.
Sociodemographic and clinical characteristics of participants according to Plasmodium falciparum infection status
| Characteristic | PCR result | P-value | |
|---|---|---|---|
| Negative | Positive | ||
| n = 92 | n = 180 | ||
| Median age (IQR) (years) | 24 (10–42) | 18 (8–28) | < 0.01 |
| Age distribution (years), no. (%) | |||
| < 5 | 13 (14) | 25 (14) | 0.96 |
| 5–15 | 16 (17) | 52 (29) | 0.04 |
| 16–29 | 30 (33) | 59 (33) | 0.98 |
| 30–39 | 8 (9) | 22 (12) | 0.38 |
| ≥ 40 | 25 (27) | 22 (12) | < 0.01 |
| Female sex, no. (%) | 50 (56) | 92 (56) | 0.95 |
| Educational attainment, no. (%) | |||
| Primary | 31 (34) | 84 (47) | 0.04 |
| Secondary | 44 (48) | 41 (23) | < 0.01 |
| Tertiary | 2 (2) | 2 (1) | 0.49 |
| Occupation, no. (%) | |||
| Farmer | 42 (46) | 94 (53) | 0.30 |
| Fisher | 8 (9) | 16 (9) | 0.96 |
| Laborer | 4 (4) | 10 (6) | 0.67 |
| Merchant | 9 (10) | 11 (6) | 0.27 |
| Office profession | 13 (14) | 13 (7) | 0.07 |
| None | 4 (4) | 19 (11) | 0.08 |
| Other | 11 (12) | 15 (8) | 0.34 |
| Wall type, no. (%) | 0.15 | ||
| Straw-and-pole | 49 (53) | 112 (62) | |
| Cement or brick | 43 (47) | 69 (38) | |
| Roof type, no. (%) | 0.03 | ||
| Thatch | 52 (57) | 126 (70) | |
| Metal | 40 (43) | 54 (30) | |
| Mean monthly income, Zambian kwacha, (SD) | 400 (150–1,000) | 200 (100–500) | < 0.01 |
| Bed net use, no. (%) | 80 (87) | 153 (86) | 0.74 |
| Indoor residual spraying in prior 6 months, no. (%) | 53 (58) | 99 (55) | 0.68 |
| Mean no. of occupants (SD) | 6.2 (2.9) | 6.2 (2.5) | 0.94 |
| Mean no. of rooms (SD) | 4.4 (1.6) | 4.4 (1.7) | 0.99 |
| Rapid diagnostic test positive, no. (%) | 8 (9) | 121 (67) | – |
| Blood smear positive, no. (%) | 0 (0) | 93 (52) | – |
| Gametocytemia, no. (%) | 0 (0) | 7 (4) | – |
| Median parasites per μL (IQR) | 17,500 | 84,960 (15,200–402,210) | – |
IQR = interquartile range. P-values were computed by Student’s t-test for continuous variables or Pearson’s χ2 test for dichotomous variables.
Association of participant characteristics with malaria.
The highest proportion of cases occurred in school-aged children, and the lowest proportion in adults older than 40 years. Adjusted for sex, prior IRS, ITN usage, and house features, school-aged children had twice the odds of infection relative to other age-groups (odds ratio [OR]: 1.9; 95% CI: 0.1–3.7, P = 0.05). Adjusted for the same conditions, adults older than 40 years had 67% reduced odds compared with other age-groups (OR: 0.33; 95% CI: 0.17–0.67, P = 0.002).
Lower income and educational attainment were also associated with greater odds of malaria (OR: 2.2; 95% CI: 1.3–3.9, P = 0.005 for income less than USD 2/day, OR: 3.2; 95% CI: 1.9–5.6, P < 0.001 for less than secondary-level education). Patient sex and ITN use were not associated with malaria.
Association of house features with malaria.
Residing in a house with a thatch roof was associated with 2.6 times greater odds of malaria than living in a house with a metal roof, including after adjustment for wall type, number of rooms, number of co-occupants, and recent IRS (OR: 2.6; 95% CI: 1.0–6.3, P = 0.04; Table 3, Figure 3). Compared with all other house types, residing in a house with the highest quality construction materials—metal roof and brick walls—was associated with 62% lower odds of malaria (OR: 0.38; 95% CI: 020–0.73, P = 0.004). Wall construction alone, number of rooms, number of occupants, or prior IRS was not associated with malaria in either unadjusted or adjusted models. Mediation analyses showed that roof type accounted for a large portion of the association between household income and malaria (24%, 95% CI: 14–82), and household income in turn accounted for 11% (95% CI: 8–19) of the association of education with malaria (Figure 2).
Unadjusted and adjusted ORs for Plasmodium falciparum infection according to house structure
| Unadjusted | Adjusted | |||||
|---|---|---|---|---|---|---|
| Characteristic | OR | 95% CI | P-value | OR | 95% CI | P-value |
| Wall type | ||||||
| Straw-and-pole | 1.45 | 0.87–2.40 | 0.16 | 0.90 | 0.39–2.07 | 0.81 |
| Cement or brick | Ref | – | – | Ref | – | – |
| Roof type | ||||||
| Thatch | 1.79 | 1.07–3.02 | 0.03 | 2.56 | 1.03–6.31 | 0.04 |
| Metal | Ref | – | – | Ref | – | – |
| Overall house type | ||||||
| Brick wall and metal roof | 0.56 | 0.33–0.94 | 0.03 | 0.38 | 0.20–0.73 | < 0.01 |
| Other | Ref | – | – | Ref | – | – |
OR = odds ratio. Estimated from logistic regression models. The adjusted model included wall type, roof type, number of rooms, number of cohabitants, and prior indoor residual spraying.
Malaria prevalence by house construction type. (A) shows comparison by roof construction (T = thatch and M = metal), (B) shows comparison by wall material (S = straw-and-pole and B = brick or cement block), and (C) shows results stratified by combinations of roof and wall types. Among patients presenting to rural health clinics with fever, living in a house with a thatch roof was associated with increased odds of malaria (adjusted OR: 2.6; 95% CI: 1.0–6.3, P = 0.04 denoted by asterisk). Error bars represent 95% CIs.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Malaria prevalence by house construction type. (A) shows comparison by roof construction (T = thatch and M = metal), (B) shows comparison by wall material (S = straw-and-pole and B = brick or cement block), and (C) shows results stratified by combinations of roof and wall types. Among patients presenting to rural health clinics with fever, living in a house with a thatch roof was associated with increased odds of malaria (adjusted OR: 2.6; 95% CI: 1.0–6.3, P = 0.04 denoted by asterisk). Error bars represent 95% CIs.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
Malaria prevalence by house construction type. (A) shows comparison by roof construction (T = thatch and M = metal), (B) shows comparison by wall material (S = straw-and-pole and B = brick or cement block), and (C) shows results stratified by combinations of roof and wall types. Among patients presenting to rural health clinics with fever, living in a house with a thatch roof was associated with increased odds of malaria (adjusted OR: 2.6; 95% CI: 1.0–6.3, P = 0.04 denoted by asterisk). Error bars represent 95% CIs.
Citation: The American Journal of Tropical Medicine and Hygiene 104, 6; 10.4269/ajtmh.20-1378
DISCUSSION
This cross-sectional study of patients presenting with fever to rural health centers in a highly malarious region of Zambia identified house structure as a risk factor for P. falciparum infection. Residing in a house with a thatched roof nearly tripled the odds of malaria compared with living in a house with a metal roof, whereas residing in a house with brick walls and a corrugated metal roof more than halved the odds compared with other house types. Higher income and greater educational attainment were also associated with lower odds of malaria. Mediation analyses suggested that roof type explained one-fourth of the association between household income and malaria, and household income explained one-tenth of the relationship between education and malaria. Notably, IRS and ITN use were not associated with malaria risk despite local investments in these vector control measures. Together, these findings attest to the significance of house structure to malaria risk: even in this population of individuals with a remarkably high pretest probability of malaria, differences in house structure were independently able to discern those who tested positive from those who tested negative. These findings also may explain, in part, the recalcitrance of malaria despite control efforts in this region and similar areas. The highest risk house feature, thatch roof, was also the most prevalent (65%).
Housing is a known, modifiable risk factor for malaria, but the evidence is equivocal with regard to the relative contributions of wall construction or roof construction.3,43–50 The present study suggests that in certain settings, roof construction and its attendant factors play a greater role in modulating malaria risk than wall construction. Concealed crannies in thatch roofs that foster higher mosquito densities,51 relatively faster parasite development and prolonged indoor dwelling due to differences in indoor temperature across different roof materials,16,52,53 ease of entry via eaves,14,54 and hindrances to bed net use (related structural features and effect on ambient environment)19,55,56 are possible mechanisms of the observed associations between thatch roofs and malaria risk.
The protective association between higher household income and malaria is mediated by several factors apart from housing. Higher income has been shown to be associated with greater uptake of malaria preventive measures such as ITN ownership and retreatment.25,27,29,30 When they fall ill, wealthier individuals are more likely to seek care and less likely to use leftover drugs than their poorer counterparts.57,58 The role of education is similarly multifarious and its association with malaria risk previously shown.23,59–61 Studies found a positive correlation of education with malaria knowledge and ITN ownership,62,63 as well as with health facility use and whether biomedical or ethnomedical care was sought.62,64
Our other results agree with previously reported findings in the literature. Higher prevalence of malaria in school-aged children than in other age groups is also seen in similar settings.65 The protective contribution of older age is explained by acquired immunity over time with repeat infections (premunition).66 The imbalance toward female adult patients (66% women versus 34% men, χ2 test, P < 0.001) suggests underuse of health clinics by men, a common experience in similar subsistence economic settings.67 The absence of a protective effect of IRS is consistent with a contemporaneous report of only marginal impact of IRS on malaria in the study area, and echoes findings of the landmark Garki Project.40,68 Outdoor transmission, incomplete spray coverage in the community, inaccurate timing of IRS in relation to malaria vector density, poor residual efficacy, or mosquito resting behaviors that avoid sprayed surfaces are among factors that could explain its low efficacy in this setting. Similarly, ITN use was not significantly associated with malaria in our sample. We suspect this is explained by the small sample size, reporting bias, and ascertainment bias due to the nature of the study population. The study population consisted of febrile clinic patients residing in a highly malarious area who therefore had a high pretest probability of malaria, and hence may represent a group of people less likely to receive a benefit from ITN use than their nonfebrile, nonclinical counterparts.
This was an observational study using cross-sectional data from rural health centers in a high malaria-transmission area Zambia and therefore bears inherent limitations in its ability to demonstrate causation and generalizability to other populations. However, the specific protective mechanisms of housing against malaria are plausible and well described in previous studies in similar settings, including recent cluster-randomized trials that examine window and eave screening and construction with alternative materials.4 This is a secondary analysis of data collected during the course of a validation study of a diagnostic test; data collection was not optimized for in-depth study of housing features. We lacked data on the presence or absence of eaves, ceilings, windows, and other potentially important features. Similarly, an ideal study of malaria and housing would incorporate entomological and geospatial data, which were not collected given the clinical scope of the parent study. The definition of malaria used herein subsumes clinical malaria but may have included incidental parasitemia occurring in the presence of non-malarial causes of fever. However, parasitemia without clinical malaria (sub-patent malaria and chronic malaria) is itself a relevant outcome to the study, given the interest in examining associations with P. falciparum infection and house structure.
Malaria is foremost a disease of rural poverty with a complex array of predisposing factors that include environmental, vector bionomic, and sociodemographic.22 The association between housing and malaria encompasses direct and indirect effects, from directly impeding mosquito entry to indirectly influencing IRS durability and bed net usage.13–16,19,20 House structure tracks predictably with income, occupation, and education, each of which entail downstream factors ranging from health-seeking behavior to access to health care and purchasing power.69 House structure appeared to account for a significant fraction (24%) of the association between income and malaria, and household income in turn explained a fair portion (11%) of the association of education with malaria. Understanding the extent of these associations can help guide malaria control policies and direct resources to those at greatest risk.
CONCLUSION
For over a century, housing modifications that reduce exposure to mosquito vectors have been recognized as a potential ward against malaria. Alongside IRS and ITNs, housing improvements represent an effective yet underused tool for vector control.3,4 Current Zambian housing regulations set standards for ventilation, overcrowding, and other public safety elements but do not take advantage of the opportunity to promote malaria control through evidence-based housing guidelines.70 Housing-directed initiatives, independent of other malaria control or economic development policies, are predicted to reduce malaria transmission in northern Zambia and similar high-transmission settings. Formulation of local and national housing standards with an eye toward vector control could help advance malaria control.
Acknowledgments:
We thank the patients and staff of Kashikishi and Nchelenge Rural Health Centers. We extend our gratitude to the Tropical Diseases Research Centre management and laboratory staff for their institutional and technical support.
REFERENCES
- 1.↑
Lane C, 1931. Housing and Malaria: A Critical Summary of the Literature Dealing with This Subject. Geneva, Switzerland: League of Nations, Health Organisation.
- 2.
Celli A, 1900. The new prophylaxis against malaria in Lazio. Lancet 156: 1603–1606.
- 3.↑
Tusting LS, Ippolito MM, Willey BA, Kleinschmidt I, Dorsey G, Gosling RD, Lindsay SW, 2015. The evidence for improving housing to reduce malaria: a systematic review and meta-analysis. Malar J 14: 209.
- 4.↑
Furnival-Adams J, Olanga EA, Napier M, Garner P, 2020. House modifications for preventing malaria. Cochrane Database Syst Rev 10: CD013398.
- 5.↑
Mukonka VM et al. 2014. High burden of malaria following scale-up of control interventions in Nchelenge district, Luapula province, Zambia. Malar J 13: 153.
- 6.↑
Eisele TP et al. 2016. Short-term impact of mass drug administration with dihydroartemisinin plus piperaquine on malaria in southern province Zambia: a cluster-randomized controlled trial. J Infect Dis 214: 1831–1839.
- 7.
Moss WJ et al. 2015. Malaria epidemiology and control within the international centers of excellence for malaria research. Am J Trop Med Hyg 93 (Suppl 3): 5–15.
- 8.
Chanda E, Kamuliwo M, Steketee RW, Macdonald MB, Babaniyi O, Mukonka VM, 2012. An overview of the malaria control programme in Zambia. ISRN Prev Med 2013: 495037.
- 9.
Chizema-Kawesha E, Miller JM, Steketee RW, Mukonka VM, Mukuka C, Mohamed AD, Miti SK, Campbell CC, 2010. Scaling up malaria control in Zambia: progress and impact 2005–2008. Am J Trop Med Hyg 83: 480–488.
- 10.
Kent RJ, Thuma PE, Mharakurwa S, Norris DE, 2007. Seasonality, blood feeding behavior, and transmission of Plasmodium falciparum by Anopheles arabiensis after an extended drought in southern Zambia. Am J Trop Med Hyg 76: 267–274.
- 11.↑
Ippolito MM, Searle KM, Hamapumbu H, Shields TM, Stevenson JC, Thuma PE, Moss WJ; For The Southern Africa International Center Of Excellence For Malaria Research, 2017. House structure is associated with Plasmodium falciparum infection in a low-transmission setting in southern Zambia. Am J Trop Med Hyg 97: 1561–1567.
- 12.↑
Paaijmans KP, Thomas MB, 2011. The influence of mosquito resting behaviour and associated microclimate for malaria risk. Malar J 10: 183.
- 13.↑
Kirby MJ, West P, Green C, Jasseh M, Lindsay SW, 2008. Risk factors for house-entry by culicine mosquitoes in a rural town and satellite villages in the Gambia. Parasit Vectors 1: 41.
- 14.↑
Lindsay SW, Snow RW, 1988. The trouble with eaves; house entry by vectors of malaria. Trans R Soc Trop Med Hyg 82: 645–646.
- 15.↑
Njie M, Dilger E, Lindsay SW, Kirby MJ, 2009. Importance of eaves to house entry by anopheline, but not culicine, mosquitoes. J Med Entomol 46: 505–510.
- 16.↑
Lindsay SW et al. 2019. Reduced mosquito survival in metal-roof houses may contribute to a decline in malaria transmission in sub-Saharan Africa. Sci Rep 9: 7770.
- 17.↑
Jatta E, Jawara M, Bradley J, Jeffries D, Kandeh B, Knudsen JB, Wilson AL, Pinder M, D'Alessandro U, Lindsay SW, 2018. How house design affects malaria mosquito density, temperature, and relative humidity: an experimental study in rural Gambia. Lancet Planet Health 2: e498–e508.
- 18.↑
Lahondere C, Lazzari CR, 2012. Mosquitoes cool down during blood feeding to avoid overheating. Curr Biol 22: 40–45.
- 19.↑
Iwashita H, Dida G, Futami K, Kaneko S, Horio M, Kawada H, Maekawa Y, Aoki Y, Minakawa N, 2010. Sleeping arrangement and house structure affect bed net use in villages along Lake Victoria. Malar J 9: 176.
- 20.↑
Correa A, Galardo AKR, Lima LA, Câmara DCP, Müller JN, Barroso JFS, Lapouble OMM, Rodovalho CM, Ribeiro KAN, Lima JBP, 2019. Efficacy of insecticides used in indoor residual spraying for malaria control: an experimental trial on various surfaces in a “test house”. Malar J 18: 345.
- 21.↑
Desalegn Z, Wegayehu T, Massebo F, 2018. Wall-type and indoor residual spraying application quality affect the residual efficacy of indoor residual spray against wild malaria vector in southwest Ethiopia. Malar J 17: 300.
- 22.↑
Tusting LS, Willey B, Lucas H, Thompson J, Kafy HT, Smith R, Lindsay SW, 2013. Socioeconomic development as an intervention against malaria: a systematic review and meta-analysis. Lancet 382: 963–972.
- 23.↑
Worrall E, Basu S, Hanson K, 2005. Is malaria a disease of poverty? A review of the literature. Trop Med Int Health 10: 1047–1059.
- 25.↑
Ahmed SM, Haque R, Haque U, Hossain A, 2009. Knowledge on the transmission, prevention and treatment of malaria among two endemic populations of Bangladesh and their health-seeking behaviour. Malar J 8: 173.
- 26.
Gingrich CD, Hanson K, Marchant T, Mulligan JA, Mponda H, 2011. Price subsidies and the market for mosquito nets in developing countries: a study of Tanzania’s discount voucher scheme. Soc Sci Med 73: 160–168.
- 27.↑
Hanson K, Worrall E, 2002. Social Marketing of Insecticide Treated Nets – Phase 2 (SMITN2), Tanzania: End of Project Household Survey Analysis. London, United Kingdom: Malaria Consortium, 43.
- 28.
Matovu F, Goodman C, Wiseman V, Mwengee W, 2009. How equitable is bed net ownership and utilisation in Tanzania? A practical application of the principles of horizontal and vertical equity. Malar J 8: 109.
- 29.↑
Rashed S, Johnson H, Dongier P, Moreau R, Lee C, Lambert J, Schaefer C, 2000. Economic impact of febrile morbidity and use of permethrin-impregnated bed nets in a malarious area II. Determinants of febrile episodes and the cost of their treatment and malaria prevention. Am J Trop Med Hyg 62: 181–186.
- 30.↑
Ziba C, Slutsker L, Chitsulo L, Steketee RW, 1994. Use of malaria prevention measures in Malawian households. Trop Med Parasitol 45: 70–73.
- 31.↑
Barat LM, Palmer N, Basu S, Worrall E, Hanson K, Mills A, 2004. Do malaria control interventions reach the poor? A view through the equity lens. Am J Trop Med Hyg 71 (Suppl 2): 174–178.
- 32.↑
Saaka M, Oosthuizen J, Beatty S, 2009. Effect of joint iron and zinc supplementation on malarial infection and anaemia. East Afr J Public Health 6: 55–62.
- 34.↑
Imai K, Keele L, Tingley D, 2010. A general approach to causal mediation analysis. Psychol Methods 15: 309–334.
- 35.↑
Handema R, Daka V, Chileshe J, Nambozi M, Kasongo W, 2018. Field evaluation of the illipro-10 illumigene malaria diagnostic by amplification of plasmodium spp. DNA. Am J Trop Med Hyg 99 (Suppl 4): 91.
- 36.↑
Pinchoff J, Chaponda M, Shields TM, Sichivula J, Muleba M, Mulenga M, Kobayashi T, Curriero FC, Moss WJ; Southern Africa International Centers of Excellence for Malaria Research, 2016. Individual and household level risk factors associated with malaria in Nchelenge District, a region with perennial transmission: a serial cross-sectional study from 2012 to 2015. PLoS One 11: e0156717.
- 37.↑
Das S, Muleba M, Stevenson JC, Norris DE, 2016. Habitat partitioning of malaria vectors in Nchelenge district, Zambia. Am J Trop Med Hyg 94: 1234–1244.
- 38.↑
Stevenson JC et al. 2016. Spatio-temporal heterogeneity of malaria vectors in northern Zambia: implications for vector control. Parasit Vectors 9: 510.
- 39.↑
Hast MA et al. 2019. The impact of 3 years of targeted indoor residual spraying with pirimiphos-methyl on malaria parasite prevalence in a high-transmission area of northern Zambia. Am J Epidemiol 188: 2120–2130.
- 40.↑
Ippolito MM et al. 2018. Risk factors for mortality in children hospitalized with severe malaria in northern Zambia: a retrospective case-control study. Am J Trop Med Hyg 98: 1699–1704.
- 41.↑
Kamau E, Tolbert LS, Kortepeter L, Pratt M, Nyakoe N, Muringo L, Ogutu B, Waitumbi JN, Ockenhouse CF, 2011. Development of a highly sensitive genus-specific quantitative reverse transcriptase real-time PCR assay for detection and quantitation of Plasmodium by amplifying RNA and DNA of the 18S rRNA genes. J Clin Microbiol 49: 2946–2953.
- 42.↑
Taylor BJ, Lanke K, Banman SL, Morlais I, Morin MJ, Bousema T, Rijpma SR, Yanow SK, 2017. A direct from blood reverse transcriptase polymerase chain reaction assay for monitoring falciparum malaria parasite transmission in elimination settings. Am J Trop Med Hyg 97: 533–543.
- 43.↑
Bousema T et al. 2010. Identification of hot spots of malaria transmission for targeted malaria control. J Infect Dis 201: 1764–1774.
- 44.
Mmbando BP, Kamugisha ML, Lusingu JP, Francis F, Ishengoma DS, Theander TG, Lemnge MM, Scheike TH, 2011. Spatial variation and socio-economic determinants of Plasmodium falciparum infection in northeastern Tanzania. Malar J 10: 145.
- 45.
Ouma P, van Eijk AM, Hamel MJ, Parise M, Ayisi JG, Otieno K, Kager PA, Slutsker L, 2007. Malaria and anaemia among pregnant women at first antenatal clinic visit in Kisumu, western Kenya. Trop Med Int Health 12: 1515–1523.
- 46.
Rulisa S, Kateera F, Bizimana JP, Agaba S, Dukuzumuremyi J, Baas L, de Dieu Harelimana J, Mens PF, Boer KR, de Vries PJ, 2013. Malaria prevalence, spatial clustering and risk factors in a low endemic area of eastern Rwanda: a cross sectional study. PLoS One 8: e69443.
- 47.
Sintasath DM, Ghebremeskel T, Lynch M, Kleinau E, Bretas G, Shililu J, Brantly E, Graves PM, Beier JC, 2005. Malaria prevalence and associated risk factors in Eritrea. Am J Trop Med Hyg 72: 682–687.
- 48.
Temu EA, Coleman M, Abilio AP, Kleinschmidt I, 2012. High prevalence of malaria in Zambezia, Mozambique: the protective effect of IRS versus increased risks due to pig-keeping and house construction. PLoS One 7: e31409.
- 49.
Wanzirah H et al. 2015. Mind the gap: house structure and the risk of malaria in Uganda. PLoS One 10: e0117396.
- 50.↑
Ye Y, Hoshen M, Louis V, Seraphin S, Traore I, Sauerborn R, 2006. Housing conditions and Plasmodium falciparum infection: protective effect of iron-sheet roofed houses. Malar J 5: 8.
- 51.↑
Ondiba IM, Oyieke FA, Ong’amo GO, Olumula MM, Nyamongo IK, Estambale BBA, 2018. Malaria vector abundance is associated with house structures in Baringo county, Kenya. PLoS One 13: e0198970.
- 52.↑
Shapiro LLM, Whitehead SA, Thomas MB, 2017. Quantifying the effects of temperature on mosquito and parasite traits that determine the transmission potential of human malaria. PLoS Biol 15: e2003489.
- 53.↑
Murdock CC, Sternberg ED, Thomas MB, 2016. Malaria transmission potential could be reduced with current and future climate change. Sci Rep 6: 27771.
- 54.↑
Animut A, Balkew M, Lindtjorn B, 2013. Impact of housing condition on indoor-biting and indoor-resting Anopheles arabiensis density in a highland area, central Ethiopia. Malar J 12: 393.
- 55.↑
Toe LP et al. 2009. Decreased motivation in the use of insecticide-treated nets in a malaria endemic area in Burkina Faso. Malar J 8: 175.
- 56.↑
von Seidlein L, Ikonomidis K, Bruun R, Jawara M, Pinder M, Knols BG, Knudsen JB, 2012. Airflow attenuation and bed net utilization: observations from Africa and Asia. Malar J 11: 200.
- 57.↑
Biritwum RB, Welbeck J, Barnish G, 2000. Incidence and management of malaria in two communities of different socio—economic level, in Accra, Ghana. Ann Trop Med Parasitol 94: 771–778.
- 58.↑
Mugisha F, Kouyate B, Gbangou A, Sauerborn R, 2002. Examining out ‐ of ‐ pocket expenditure on health care in Nouna, Burkina Faso: implications for health policy. Trop Med Int Health 7: 187–196.
- 59.↑
Kreuels B, Kobbe R, Adjei S, Kreuzberg C, von Reden C, Bäter K, Klug S, Busch W, Adjei O, May J, 2008. Spatial variation of malaria incidence in young children from a geographically homogeneous area with high endemicity. J Infect Dis 197: 85–93.
- 60.
Pullan RL, Kabatereine NB, Bukirwa H, Staedke SG, Brooker S, 2011. Heterogeneities and consequences of Plasmodium species and hookworm coinfection: a population based study in Uganda. J Infect Dis 203: 406–417.
- 61.↑
Slutsker L et al. 1996. Malaria infection in infancy in rural Malawi. Am J Trop Med Hyg 55 (Suppl 1): 71–76.
- 62.↑
Fawole OI, Onadeko MO, 2001. Knowledge and home management of malaria fever by mothers and care givers of under five children. West Afr J Med 20: 152–157.
- 63.↑
Nuwaha F, 2001. Factors influencing the use of bed nets in Mbarara municipality of Uganda. Am J Trop Med Hyg 65: 877–882.
- 64.↑
Kaona F, Siajunza M, Manyando C, Khondowe S, Ngoma G, 2000. Utilisation of malarial drugs at a household level: results from a KAP study in Choma, southern province and Mporokoso, northern province of Zambia. Cent Afr J Med 46: 268–270.
- 65.↑
Walldorf JA et al. 2015. School-age children are a reservoir of malaria infection in Malawi. PLoS One 10: e0134061.
- 66.↑
Sergent E, Parrot LL, 1935. Immunité, la Prémunition et la Résistance Innée. Archives de l'Institut Pasteur d'Algérie, TXIII(279).
- 67.↑
Ippolito MM, Chary A, Daniel M, Barnoya J, Monroe A, Eakin M, 2017. Expectations of health care quality among rural Maya villagers in Solola department, Guatemala: a qualitative analysis. Int J Equity Health 16: 51.
- 68.↑
Molineaux L, Grammiccia G, 1980. The Garki Project: Research on the Epidemiology and Control of Malaria in the Sudan Savanna of West Africa. Geneva, Switzerland: World Health Organization.
- 70.↑
Government of the Republic of Zambia, 2020. Zambia National Public Health Act. Act IX § 67. Lusaka, Zambia: National Assembly of Zambia.