What if a nutrient deficiency was widespread across rural communities, yet still changed sharply from one district to another? This pilot survey maps selenium status in Zimbabwean children and women—and finds an especially serious risk among children. Selenium deficiency is increasingly recognized as a public health problem in sub-Saharan Africa.

That sets up the central question: how common is it, and how does it vary geographically in rural Zimbabwe? The study frames selenium deficiency as an increasingly recognized public health problem in sub-Saharan Africa, emphasizing its importance for population health.

The current cross-sectional study assessed the prevalence and geospatial patterns of selenium deficiency among children aged six to fifty-nine months and women of fifteen to forty-nine years old, selected by simple random sampling in rural Zimbabwe. The survey covered Murewa, Shamva, and Mutasa districts, while venous blood samples were collected and stored according to World Health Organization guidelines.

Plasma selenium concentration was determined by inductively coupled plasma-mass spectrometry, providing the biomarker used throughout the analysis. The cross-sectional biomarker survey included eight hundred thirty-one women of reproductive age and seven hundred forty-one children aged six to fifty-nine months, and it was conducted between October twenty twenty-one and January twenty twenty-two.

The survey covered three rural districts: Murewa, Shamva, and Mutasa in Zimbabwe. These districts were chosen because of high stunting rates and low dietary diversity, with only fifteen to twenty-nine point nine percent of children receiving the minimum number of required food groups.

The sampling design was nested at the level of the National Demographic Health Survey sampling approach. Figure one maps the sampling locations of eligible households across Shamva, Murewa, and Mutasa districts in rural Zimbabwe. Black points mark enumeration areas, while outlined regions show district boundaries; the inset places these districts within Zimbabwe.

The figure matters because it makes the study’s geographic coverage visible, complementing the sampling design in which thirty enumeration areas were selected per district and ten households were randomly selected from each area. Replicate plasma samples were shipped on dry ice to the University of Nottingham in the United Kingdom for quantification of plasma selenium concentration using inductively coupled plasma-mass spectrometry.

The limit of detection was measured as three times the standard deviation of ten operational blanks, and the limit of quantification was calculated as ten times this standard deviation. The limit of detection and limit of quantification were zero point zero two nine and zero point zero nine six micrograms per liter, respectively.

Accuracy was verified using two reference materials, and average selenium recovery across forty analytical batches was ninety-nine and one hundred two percent for the two materials. Prevalence of selenium deficiency was compared against plasma selenium thresholds of seventy micrograms per liter, and against thresholds previously indicated as optimal activities of SELENOP, GPX3, and IDI.

Exploratory analysis used summary statistics, plots, and quantile-quantile plots to check for outliers and data normality, while classified post plots were used to check for spatial trends. Ordinary kriging was then used to predict plasma selenium concentration at unsampled households for women of reproductive age and children in the three districts.

Cross-validation omitted each observed data point in turn, predicted its value by ordinary kriging, and assessed model fit using the median standardized squared prediction error. The exponential and spherical models were both plausible. Table one summarizes the study participants: children aged six to fifty-nine months and women aged fifteen to forty-nine years, distributed across Murewa, Shamva, and Mutasa.

Among the children, the male and female counts were 368 and 373, respectively, with a median age of 28.8 months. The women had a median age of 30 years. This matters because it shows the demographic composition and district coverage of the sample used to assess selenium deficiency.

Selenium deficiency was highly prevalent in both women and children in all three rural districts. Children were more likely to be deficient than women: ninety-four point six percent of children, compared with sixty-nine point eight percent of women.

The overall median plasma selenium concentrations were sixty-one point two micrograms per liter for women and forty point five micrograms per liter for children. The sample also showed evidence of severe selenium deficiency risk, with concentrations as low as nine point four one micrograms per liter in children and ten point two zero micrograms per liter as the minimum in women of reproductive age.

Plasma selenium concentrations differed between districts, ranging from a median of thirty-six point five micrograms per liter in Murewa to seventy-one point seven micrograms per liter in Mutasa. In Mutasa, the demographic groups also differed substantially: the median was seventy-one point seven micrograms per liter in women of reproductive age and forty-seven point two in children.

Table two summarizes selenium status across Murewa, Shamva, and Mutasa for children aged six to fifty-nine months and women aged fifteen to forty-nine years. Overall median plasma selenium was forty point five micrograms per liter in children and sixty-one point two in women, while the table reports deficiency prevalence using plasma selenium, SELENOP, GPX3, and IDI thresholds.

It matters because deficiency was widespread across all districts, with overall plasma-selenium deficiency reported at ninety-six point four percent for children and sixty-nine point eight percent for women. The variograms for both children and women of reproductive age do not show large marked spatial trends, and selenium status was reasonably homogeneous within each district.

At the same time, there was substantial inter-district variation across a snapshot of three areas of approximately fifty by fifty kilometers. The spatial analysis makes hot spots of selenium deficiency visible: lighter purple shades indicate low plasma selenium concentrations, while darker shades indicate high concentrations.

Murewa had the lowest plasma selenium concentrations in both groups, while Mutasa recorded the highest values. Table three summarizes plasma selenium distributions across Murewa, Shamva, and Mutasa for children aged six to fifty-nine months and women aged fifteen to forty-nine years.

It reports sample counts, means, medians, quartiles, variance, standard deviation, and octile skewness; the counts represent sampling points with valid geographic coordinates. These descriptive statistics show district-level variation and provide the numerical basis for the accompanying geospatial analysis of selenium status.

Table four evaluates exponential and spherical covariance models using cross-validation, reported as median standardized squared prediction error, or SSPE, across Murewa, Shamva, and Mutasa. For children aged six to fifty-nine months, the values range from zero point three two three to zero point four five five; for women of reproductive age, they range from zero point two nine five to zero point four one six.

The authors note that a valid variogram has an expected median SSPE of zero point four five five, so these results inform covariance-model selection while also showing the prediction uncertainty considered in the spatial analysis. Figure two maps kriging predictions of plasma selenium across Shamva, Murewa, and Mutasa, separately for children aged six to fifty-nine months and women of reproductive age.

Purple shading represents predicted concentration, while its spatial pattern shows relatively homogeneous selenium status within districts but substantial variation between districts. The authors use these maps to visualize potential hotspots of selenium deficiency, with reported probabilities of concentrations below functional thresholds especially high in Murewa and Shamva.

For unsampled children, the average probability of plasma selenium falling below the threshold for optimal IDI activity was zero point ninety-eight in Murewa, zero point ninety-eight in Shamva, and zero point nine in Mutasa. For unsampled women of reproductive age, the average probability of falling below the threshold for optimal GPX3 activity was zero point ninety-seven in Murewa, zero point ninety-eight in Shamva, and zero point seventy-three in Mutasa.

The verbal scale shows the likelihood of being below a threshold, and the brown color tone reflects a higher likelihood of selenium deficiency. Children aged six to fifty-nine months residing in Murewa had the highest likelihood of being selenium deficient.

Figure three maps ordinary-kriging variance for predicted plasma selenium concentrations in children aged six to fifty-nine months and women of reproductive age across Shamva, Murewa, and Mutasa. Darker and lighter shading represent different uncertainty levels at unsampled locations, rather than selenium concentration itself.

The authors report high kriging variances in both groups: average values range from one hundred forty-two to one hundred forty-four in children and from two hundred seventy-seven to three hundred ninety-one in women, with Shamva showing slightly lower variances. Figure four maps the probability that plasma selenium falls below functional thresholds: below sixty-four point eight micrograms per liter for IDI in children aged six to fifty-nine months, and below eighty-four point nine micrograms per liter for GPX3 in women of reproductive age.

Across Shamva, Murewa, and Mutasa, the authors report very high average probabilities for unsampled children—zero point nine eight, zero point nine eight, and zero point nine—and for women, zero point nine seven, zero point nine eight, and zero point seven three. This matters because it indicates geographically widespread risk of selenium insufficiency.

The median plasma selenium concentrations in women were similar to those established in other African countries, including Malawi and Ethiopia. Children’s median concentrations were low compared with China, a known selenium-deficient country.

Globally, average plasma selenium concentrations fall within eighty to one hundred twenty micrograms per liter, showing the magnitude of selenium deficiency risk among the Zimbabwean population. Different adequacy thresholds reflect the expression of essential selenoproteins: SELENOP, GPX3, and IDI are maximal above one hundred, eighty-four point nine, and sixty-four point eight micrograms per liter, respectively.

The high selenium deficiency prevalence is consistent with similarities in geography, diet, and socioeconomic conditions between Zimbabwe and Malawi. In both countries, rural populations rely on highly localized agriculture and food systems, with dietary intakes supplied through subsistence production and locally based food purchases.

When localized food systems occur in low selenium environments, the entry of selenium into diets is limited. Zimbabwe also has limited access to freshwater fish, an essential source of selenium and other micronutrients. The findings indicate that selenium deficiency is likely widespread in Zimbabwe, creating a pressing need for surveillance in other regions and demographic groups and for a nationally representative biomarker survey.

The study is limited in geographic scope, so it cannot infer selenium status or deficiency risk across Zimbabwe, especially because selenium status is likely under strong geographic control. There is no consensus on the most appropriate threshold for indicating selenium deficiency risk.

The study measured only plasma selenium concentration as a proxy for SELENOP, GPX3, and IDI status, so it may not fully reflect selenoprotein status. Additional direct measurements of selenoproteins would complement these findings, although the study contributes to geostatistical modeling for interpolation and mapping of nutrient deficiencies.

The study showed that selenium deficiency was widespread among children aged six to fifty-nine months and women of reproductive age in the three pilot districts. Spatial controls on selenium status were driven by short-range variation, with no marked spatial pattern within districts but substantial variation between districts.

The prevalence of deficiency and the very low plasma selenium concentrations recorded in the selected districts are cause for concern. Further population-level investigations are warranted to measure the magnitude of the problem using reassessed dietary reference values and upper intake levels, build more evidence, and map strategies to reduce selenium deficiency.

Across the three pilot districts, selenium deficiency was widespread, children were more affected than women, and district-level differences were substantial. The results support broader, nationally representative biomarker surveillance before strategies are scaled up.