Rural-urban disparities in the nutritional status of younger adolescents in Tanzania.

Research on geographic differences in health focuses largely on children less than five years; little is known about adolescents-and even less regarding younger adolescents-a vulnerable group at a critical stage of the life course. Africa's rapid population growth and urbanization rates, coupled with stagnant rates of undernutrition, further indicate the need for country-specific data on rural-urban health disparities to inform development policies. This study examined rural-urban disparities in body mass index-for-age-and-sex (BAZ) and height-for-age-and-sex z-scores (HAZ) among younger adolescents in Tanzania. Participants were randomly selected adolescents aged 10-14 years (N = 1,125) residing in Kilosa (rural) and Moshi (urban) districts of Tanzania. Individual and household-level data were collected using surveys and anthropometric data was collected on all adolescents. Age, sex, household living conditions, and assets were self-reported. BAZ and HAZ were calculated using the WHO reference guide. The prevalence of undernutrition was 10.9% among rural and 5.1% among urban adolescents (p<0.001). Similarly, stunting prevalence was greater in rural (64.5%) than urban (3.1%) adolescents (p<0.001). After adjusting for covariates, rural residence was significantly and inversely associated with BAZ (B = -0.29, 95% CI: -0.52, -0.70, p = 0.01), as well as with HAZ (B = -1.79, 95% CI: -2.03, -1.54, p<0.001). Self-identified males had lower BAZ (B = -0.23, 95% CI: -0.34, -0.11, p<0.001) and HAZ (B = -0.22, 95% CI: -0.35, -0.09, p = 0.001) than self-identified female adolescents. Rural-urban disparities in nutritional status were significant and gendered. Findings confirm place of residence as a key determinant of BAZ and HAZ among younger adolescents in Tanzania. Targeted gender-sensitive interventions are needed to limit growth faltering and improve health outcomes in rural settings.

Introduction

Undernutrition is the leading cause of morbidity and mortality worldwide, contributing to an estimated 3.1 million deaths per year [1]. Existing literature across Sub-Saharan Africa (SSA) affirms an urban health advantage in relation to stunting [2-8], underweight [2,7], and body mass index (BMI) [6-8], suggesting that place of residence is an important social determinant of health outcomes. Most of the literature on geographic differences in health focuses on children under five years; little is known about adolescents, a vulnerable group at a critical stage of the life course. Emerging research on points of intervention to protect adolescent health, particularly during early adolescence, can present new policy and programmatic opportunities to promote child survival and improve adult health outcomes [1,9].

Adolescence, defined as 10–24 years of age [10], is second to infancy in linear growth velocity. This period of accelerated physical, hormonal, and cognitive development is accompanied by increased nutritional demand, which places adolescents in resource-constrained areas at greater risk for undernutrition. While research reveals the prevalence of undernutrition among adolescents in SSA ranges between 7% and 27% [3,6-8,11,12], there is a paucity of data examining adolescent health by rural versus urban residence. The few studies in this region lack consistency in findings and do not specifically address younger adolescents. While most of these studies confirmed an urban advantage in nutritional status, this study comprehensively examines age, sex, and place of residence as a determinant of adolescent nutritional health. Furthermore, this study adds to the limited knowledge of these intersections and presents disaggregated data on rural and urban adolescents in Tanzania, which aids the understanding of health risks affecting those transitioning from childhood into adulthood.

In Ethiopia, Berheto et al. (2015) found significantly lower mean BMI-for-age-and-sex z-scores (BAZ), lower mean height-for-age-and-sex z-scores (HAZ), and higher rates of stunting among rural versus urban adolescent girls [6]. Similarly, Hadley et al. (2011) reported lower weight-for-age-and-sex z-scores, BMI, HAZ, and BAZ among Ethiopian rural compared to urban adolescents after controlling for income, age, sex, and workload [7]. In Nigeria, Omigbodun et al. (2010) observed that rural adolescents had lower BAZ and were significantly more likely to be stunted and underweight than their urban peers [8]. Irenso et al. (2020) noted significantly lower HAZ and higher rates of stunting among rural compared to urban Ethiopian adolescents [3]. Lastly, in a small sample in Cameroon, Dapi et al. (2010) reported no differences in stunting among rural and urban adolescents [13].

Adolescents remain one of the most neglected, difficult to measure, and hard-to-reach populations. With rapid urban growth in SSA, a focus on adolescent health across geographic location is ever more vital to inform policy. Determinants of rural-urban health disparities can inform programmatic options to mitigate health consequences associated with migration and rapid urbanization, especially given that the nutritional status of the lowest quartiles of poor urban youth are on par with or worse than rural inhabitants [14,15].

Younger adolescents are often excluded from programs and policies that target children or young adults [16]. Few studies have examined place of residence as a determinant of nutritional health among younger adolescents and there are limited studies available to draw conclusions about urban-rural disparities in adolescent nutritional status across SSA. This study aimed to investigate rural/urban differences in BAZ and HAZ among adolescents, aged 10–14 years, living in Kilosa (rural) and Moshi (urban) districts of Tanzania. In addition to having one of the largest rural-urban gaps in malnutrition in SSA [17], 44% of Tanzania’s population is under the age of 15 years and 70% of citizens live in rural areas [18]. Our study presents disaggregated data on adolescents from Tanzania, which is needed to inform health priorities and intervention points for adolescent health.

Methods

Study population

This cross-sectional comparative study merged and collectively analyzed two independent samples of adolescents (10–14 years) from Tanzania. In 2004, Cordeiro’s study (Kilosa/rural) was conducted in rural Tanzania [11] and another study by Carlson and Earls (Moshi/urban) was conducted in an urban setting [19]. At the time of this study, Kilosa consisted of 161 registered rural villages [20] across 37 wards, and had a total population of 488,191. Moshi Urban District, the capital of the Kilimanjaro Region, comprised 15 wards totaling 143,799 people in 2002 [20].

The rural dataset was derived from a two-stage sampling plan for selection of villages and adolescents aged 10–19 years from Kilosa District, as detailed by Cordeiro et al. (2012) [11]. The Moshi/urban dataset was derived from a two-stage sampling plan for a cluster randomized control trial conducted in Moshi Urban District among adolescents aged 10–14 and their households [19]. From the combined sample, 1,237 were younger adolescents, aged 10–14 years. After eliminating 112 cases due to missing or incomplete data, the total analytical sample for this study was 1,125 adolescents aged 10–14 years.

Data collection

The survey tool was developed in English based on an extensive literature review, and translated into Swahili by language experts in Tanzania, and then back-translated into English. The questionnaires were pre-tested among a sample of non-participants in Kilosa District and Moshi prior to using it with study participants. Trained interviewers administered the structured, pre-tested questionnaires in Swahili. Content validity was conducted, ensuring that terms and meanings were understood by participants in the pretest and that terms were accurately translated from English to Swahili. Construct validity required setting a priori hypotheses of associations and analyzing the pilot data to see if these hypotheses signaled validity. For quality assurance, piloting the survey tool among adolescents and their parents/guardian in both Kilosa and Moshi resulted in no major changes in the content and construct of the survey. Surveys in both English and Swahili are included as supporting information.

Both studies used 2002 Tanzanian census data to determine household composition, and all household demographics were verified during home visits. Surveys were administered in Swahili by trained interviewers who collected demographic, dietary, health, work status, developmental, and educational data. Adolescents and their caregivers were interviewed at the adolescent’s residence or schools. Under the supervision of the research directors, height and weight were measured twice for each adolescent and averaged. Height was measured to the nearest 0.1cm with a standard stadiometer and weight was measured to the nearest 100g using calibrated UNICEF electronic scales. In the urban study, a senior research assistant repeated the physical exam measures on every fifth participant initially measured by a junior research assistant. Junior research assistants with more than two unreliable measurements were removed from the interviewing process and offered retraining.

For both studies, heads of households or guardians reported on demographics, health, and socioeconomic characteristics. To address self-reporting bias, survey instruments were carefully designed, piloted, and improved in the early stages of the study. Survey items were assessed for potential bias. Data on key variables were collected at the individual and household levels to address social desirability and selective recall bias. The main reasons for non-participation were hearing impairment, illness, nonresponse, refusal, and language barriers.

Variables

Outcome variables included BAZ; HAZ; undernutrition, defined as BAZ < -2 standard deviations (SD) of the WHO 2007 growth reference; and stunting, defined as HAZ < -2SD of the WHO 2007 growth reference. Dichotomous variables categorizing individuals as undernourished (BAZ < -2SD) and stunted (HAZ < -2SD) were used for descriptive statistics. BAZ and HAZ were entered as continuous dependent variables in multivariate linear regression analyses.

Individual-level independent variables included age, sex, orphan status, school enrollment, work status, and health index. Age was calculated using the birth date and the date of survey administration. Sex was self-reported (females, 0; males, 1). Orphan status was determined based on death of either one parent or both parents and was verified at the household level (non-orphans, 0; orphans, 1). School enrollment was reported by the adolescent and verified at the household level (not enrolled, 0; enrolled, 1). Work status reflected participation in formal and informal work activities (not working, 0; working, 1).

Health index was based on self-reported susceptibility or incidence of malaria, persistent cough, and diarrhea. The rural data included self-reported incidence of diarrhea, cough, and malaria within the past 2–3 months, while the urban data reported on adolescents’ perceived susceptibility to these illnesses. Discrepancies in health assessments across the two datasets are a limitation of this index. Health indicators were transformed into dichotomous variables for analysis, resulting in a health index score of 0 to 3 (no reported malaria, persistent cough, or diarrhea, 0; susceptibility or incidence of all three health conditions, 3).

Household level variables included place of residence, assets, and living conditions. Place of residence represented the primary predictor and was dichotomously categorized as urban (Moshi, 0) or rural (Kilosa, 1). Potential confounders included household assets and household living conditions, proxy indicators of wealth and socioeconomic status index. Household assets included ownership of a radio, bike, and/or motorbike, equally weighted and dichotomously classified. The asset index used a scale of 0–3 (poor/least affluent households, 0; most affluent households, 3). Household living conditions represented the sum of the following four indicators, each weighted equally and scored dichotomously: electricity, house flooring type (cement/stone was considered modern), source of drinking water (piped, public tap, and neighbor’s water were considered clean water sources), and access to a sanitary toilet (flush, pour flush, and improved pit latrines were considered sanitary). These indicators are correlated with human welfare and economic development. A household with a fully modern or more affluent status had a living condition score of 4 with household access to electricity, modern flooring, clean water, and a sanitary toilet; a score of ≤ 1indicates the poorest status with ≤ 1 of these resources.

Statistical analysis

SPSS Version 26 (Armonk, NY, USA, 2019) was used to analyze the data. Median values, means, standard deviations, skewness, and kurtosis were estimated for continuous variables. Missing data met the assumptions for data missing at random within the respective datasets. Due to the complexity of collecting data across rural villages, missing data was higher in Kilosa/rural than in Moshi/urban. Missing data was minimized through return interviews conducted in Kilosa/rural. Multiple imputation was applied to height and weight, indicator variables for generating the dependent variables BAZ, HAZ, undernutrition, and stunting. To retain the sample size and reflect the characteristics of the merged dataset, missing data for dependent and independent variables were retained but were excluded in analyses.

Pearson’s chi-squared test examined associations between undernutrition or stunting, health index, sociodemographic variables, and place of residence. An independent-samples t-test compared the means for BAZ and HAZ by rural/urban residence. Scatterplots, Pearson’s correlation, and univariate linear regression analyses were applied to the data. Variables with p-values ≤ 0.25 were included in multivariate linear regression models.

It was assumed that there was no first order linear autocorrelation in the multiple linear regression data given Durbin-Watson statistics, and tolerance > 0.1 and VIF < 10 for all variables in the regression models. Furthermore, assessment of normality of the residuals with normal P-P plots indicated that residuals were normally distributed. We present the findings on the association between rural/urban residence by BAZ and HAZ, respectively, after adjusting for covariates. In multivariate linear regression analyses, Model 1, with BAZ as the outcome variable, adjusted for age, sex, household assets, and household living conditions. Model 2, with HAZ as the outcome variable, adjusted for age, sex, school enrolment, household assets, and household living conditions.

Ethical considerations

Prior to participation, all adolescents provided their written assent and their parent(s) or guardian(s) provided written informed consent. The rural study was approved prior to any survey development or data collection by the Institutional Review Board at Tufts University and the Tanzania Commission for Science and Technology (COSTECH). Morogoro Regional Administration and Kilosa District Council granted field research approval. Research approval and ethical clearance for the urban study were obtained from Harvard University, the Ethical Clearance Committee of the Kilimanjaro Christian Medical College, and the Tanzanian National Institute of Medical Research. Ethics and data safety observations were provided by a Data Safety and Monitoring Board before the urban study began. Both studies were registered with COSTECH. This study was reviewed and approved by the University of Massachusetts at Amherst School of Public Health and Health Sciences Local Human Subjects Review Board before the study began.

Results

This study included 1,125 adolescents aged 10–14 years (51.4% males, 48.6% females). Thirty-seven percent (36.7%) of adolescents lived in Kilosa/rural and 63.3% lived in Moshi/urban. Significant differences were found across rural and urban settings in orphan status (8.6% rural vs. 17.7% urban, p < 0.001), school enrollment (90% rural vs. 96.5% urban, p < 0.001), and work status (51% rural vs. 60.5% urban, p = 0.002) (Table 1).

Table 1

Characteristics of adolescents (10–14 yrs) in Kilosa/rural and Moshi/urban (N = 1,125).

Characteristics	n a	% of Total	Rural (%)	Urban (%)	p b
Overall	1125	100	36.7	63.3
Age					0.002
10 years	298	26.6	24.5	27.8
11 years	220	19.6	16.9	21.2
12 years	256	22.8	21.5	23.6
13 years	217	19.3	25.7	15.7
14 years	131	11.7	11.4	11.8
Data missing	3		3	0
Sex					0.920
Male	578	51.4	51.6	51.3
Female	547	48.6	48.4	48.7
Orphan Status					<0.001
Orphan	161	14.4	8.6	17.7
Non-orphan	955	85.6	91.4	82.3
Data missing	9		8	1
School Enrollment					<0.001
Not enrolled	66	5.9	10.0	3.5
Enrolled	1054	94.1	90.0	96.5
Data missing	5		5	0
Work Status					0.002
Working	637	57.0	51.0	60.5
Not working	480	43.0	49.0	39.5
Data missing	8		7	1
Health and Nutritional Indicators	n a	% of Total	Rural (%)	Urban (%)	p b
Health Indexc					0.018
No susceptibility or incidence of malaria, persistent cough, or diarrhea (0)	463	41.6	45.7	39.2
One (1) of the following: malaria, persistent cough, or diarrhea	313	28.1	29.6	27.2
Two (2) of the following: malaria, persistent cough, or diarrhea	216	19.4	16.3	21.2
Three (3) of the following: malaria, persistent cough, or diarrhea	122	11.0	8.4	12.4
Data missing	11		8	3
Undernutritiond					<0.001
Not undernourished	1038	92.8	89.1	94.9
Undernourished	81	7.2	10.9	5.1
Data missing	6		1	5
Stuntinge					<0.001
Not stunted	833	74.4	35.5	96.9
Stunted	287	25.6	64.5	3.1
Data missing	5		2	3

a Totals may differ due to missing data on some variables. Most missing data were in the rural setting, with the exception of undernutrition and stunting.

b Tests of statistical significance are based on two-tailed Pearson χ2, p <0.05, p <0.01, p <0.001.

c Health Index defined as susceptibility to (urban) or incidence of (rural) diarrhea, malaria, or cough.

d Undernutrition defined as BMI-for-age <-2SD according to WHO 2007 reference.

e Stunting defined as height-for-age <-2SD according to WHO 2007 reference.

Fifty-four percent (54.3%) of rural compared to 60.8% of urban adolescents were susceptible to or reported an incidence of malaria, persistent cough, or diarrhea (p = 0.018) (Table 1). Significant differences were observed for undernutrition (10.9% rural versus 5.1% urban, p < 0.001) and stunting (64.5% rural versus 3.1% urban, p < 0.001) (Table 1). Rural-urban differences in ownership of household assets and living conditions are reported in Table 2.

Table 2

Household assets, housing conditions, water and sanitation in Kilosa/rural and Moshi/urban (N = 1,109).

Household Characteristics	n a	% of Total	Rural (%)	Urban (%)	p b
Overall	1109	100	35.8	64.2
Assets
Radio	875	79.4	25.8	74.2	<0.001
Bicycle	453	41.0	43.5	56.5	<0.001
Motorcycle	261	23.6	1.9	98.1	<0.001
Housing Conditions
Electricity	417	37.2	9.1	90.9	<0.001
Modern Floor	626	56.5	7.8	92.2	<0.001
Sanitary Toilets	720	66.1	1.9	98.1	<0.001
Clean Water	898	81.3	23.6	76.4	<0.001

a Totals may differ due to missing data on some variables. Except for electricity, most missing data were in the rural area: Radio (23); bicycle (20); motorcycle (20); electricity (4); modern floor (18); sanitary toilet (35); clean water (20).

b Tests of statistical significance are based on two-tailed Pearson χ2, p <0.001.

Body mass index-for-age and sex z-scores (BAZ) and height-for-age and sex z-scores (HAZ)

The mean BAZ was lower among rural (M = -0.98, SD = 0.88) than urban (M = -0.34, SD = 1.03) adolescents; t(1117) = 10.59, p < 0.001. The trajectory of median BAZ across age in urban females differed from their rural peers (data were smoothed), with the greatest rural disadvantage in BAZ observed at age 14 (Fig 1). Median BAZ of males was lower across age in rural and urban areas, with a sharper downward trajectory observed in rural males (Fig 1).

Fig 1

Median BAZ for male and female adolescents (10–14 years old) living in Moshi and Kilosa (N = 1,119).

The mean HAZ was significantly lower among rural (M = -2.29, SD = 1.11) compared to urban (M = -0.09, SD = 1.17) adolescents; t(1123) = 31.08, p < 0.001. Median HAZ declined across age (data were smoothed) by sex and place of residence (Fig 2).

Fig 2

Median HAZ for male and female adolescents (10–14 years old) living in Moshi and Kilosa (N = 1,125).

Multivariate-adjusted BAZ and HAZ by rural/urban status

After adjusting for age, sex, household assets and living conditions, an inverse association was observed between BAZ and rural residence (B = -0.29, 95% CI: -0.52, -0.07, p = 0.01). The multiple regression model produced an R = 0.13, F(5,1045) = 32.19, p < 0.001 (Table 3). The regression equation can be illustrated as follows: a 12-year-old male living in a poor rural household would have an estimated BAZ of -1.18, or be mildly undernourished, while his peer living in a poor urban household would have a BAZ of -0.88, within the normal range of the WHO standard. If the 12-year-old lived in a modern rural household, his BAZ would be -0.60. For a 12-year-old female, the pattern was similar with an estimated BAZ of -0.95 in a poor rural household compared to -0.66 in an urban poor household and -0.37 in a rural modern household (Table 4).

Table 3

Multivariate-adjusted BAZ and HAZ by rural/urban status among adolescents, 10–14 years, living in Tanzania (N = 1,125) .

BAZbN = 1050				HAZcN = 1056
Covariates	Coefficient (SE)	95% CI	p d	Covariates	Coefficient (SE)	95% CI	p d
(Intercept)	0.231 (0.29)	-0.34, 0.80	0.428	(Intercept)	2.918 (0.36)	2.21, 3.63	<0.001
Age	-0.084 (0.02)	-0.13, -0.04	<0.001	Age	-0.268 (0.02)	-0.32, -0.22	<0.001
Male	-0.226 (0.06)	-0.34, -0.11	<0.001	Male	-0.218 (0.07)	-0.35, -0.09	0.001
				School enrolled	-0.386 (0.16)	-0.69, -0.08	0.013
Assets	0.075 (0.03)	0.01, 0.14	0.015	Assets	0.057 (0.03)	-0.01, 0.12	0.101
Household living conditions index	0.119 (0.04)	0.04, 0.19	0.002	Household living conditions index	0.152 (0.04)	0.07, 0.24	<0.001
Rural residence	-0.294 (0.11)	-0.52, -0.07	0.010	Rural residence	-1.786 (0.13)	-2.03, -1.54	<0.001

Abbreviations: BAZ body mass index-for age and sex z-score; HAZ height-for-age and sex z-score.

a Models using BAZ and HAZ, normalized using the two-step data transformation process, resulted in almost identical coefficients, SE, and p values to the original data.

b R2 = 0.13; no change in r2 was observed when rural residence was removed from this model.

c R2 = 0.53; removal of rural residence from the model resulted in an r2 = 0.43, suggesting that rural residence explained 10% of the variance in the model.

d Statistical significance assessed at the p <0.05, p <0.01, p <0.001 levels.

Table 4

Estimated BAZ by rural/urban residence and household living conditions, for a reference adolescent.

Reference Adolescent	Rural or Urban	Living Conditionsa	Estimated BAZ
12-year-old male	Rural	Poor	-1.18
		Modern	-0.60
	Urban	Poor	-0.884
12-year-old female	Rural	Poor	-0.95
		Modern	-0.37
	Urban	Poor	-0.66

a Poor indicates presence ≤1 item in the household living condition index and no household assets. Modern indicates presence of all four items in the household living condition index and all three household assets.

An inverse association was observed between HAZ and rural residence, after adjusting for age, sex, school enrollment, household assets and living conditions (B = -1.79, 95% CI: -2.03, -1.54, p < 0.001) (Table 3). The multiple regression model produced an R2 = 0.53, F(6,1046) = 198.79, p < 0.001. The regression equation is illustrated by the following example: A 12-year-old school-enrolled male living in a poor rural household would be stunted with an estimated HAZ of -2.54, while his school-enrolled peer living in a poor urban household would have a normal estimated HAZ of -0.75. Even if the school-enrolled 12-year-old male lived in a modern rural household, he would be stunted with an estimated HAZ of -1.91. The pattern is similar for a 12-year-old, school-enrolled female with an estimated HAZ of -2.32 in a poor rural household compared to -0.53 in a poor urban household or -1.69 in a modern rural household (Table 5).

Table 5

Estimated HAZ by rural/urban status and household living conditions, for a reference school-enrolled adolescent.

Reference Adolescent	Rural or Urban	Living conditionsa	Estimated HAZ
12-year-old male enrolled in school	Rural	Poor	-2.536
		Modern	-1.909
	Urban	Poor	-0.884
12-year-old female enrolled in school	Rural	Poor	-0.95
		Modern	-0.37
	Urban	Poor	-0.66

a Poor indicates presence ≤1 item in the household living condition index and no household assets. Modern indicates presence of all four items in the household living condition index and all three household assets.

Discussion

The study aimed to investigate predictors of BAZ and HAZ, as well as the differences in magnitude of undernutrition and stunting, among 1,125 young adolescents living in rural and urban settings of Tanzania. Rural adolescents experienced lower mean BAZ and HAZ, and were at higher risk of undernutrition and stunting, when compared to their urban peers. Place of residence, age, and sex were significant determinants of undernutrition, with rural males from households with fewer assets and poorer living conditions being most vulnerable to undernutrition compared to their female counterparts and urban peers. Similarly, there was a higher risk of stunting among rural adolescents, particularly among the youngest adolescents, males, and those living in households with few assets. These findings are consistent with the literature across Africa showing an urban advantage in BAZ [6-8] and HAZ [3,6-8] among adolescents.

The higher risk of undernutrition and stunting in rural adolescents may be related to poorer food and health access in rural settings [4,11]. Furthermore, access to and enrollment in school is better in urban areas [21], and both adolescent and maternal education offer a protective effect against poor nutritional status [2,4,5,22]. Malnutrition has adverse implications on health, chronic diseases, and economic productivity [1]. Better health is also positively associated with SES [2,4,13-15,17], and differences in SES across geographic location are thought to contribute to growing rural-urban health disparities [4,17]. However, these disparities persist after adjusting for sociodemographic factors [4,14,23], indicating that health is influenced by many contextual factors, and, as found in our study, place of residence is an independent and salient predictor of health outcomes. Furthermore, undernourished pre-adolescents may experience further deficits in reaching their height potential during adolescence, with differential implications for male and rural youth.

Household assets reported in our study showed similar trends to national data: 79% of our households owned a radio and 41% owned a bicycle, compared to 58% and 38%, respectively [24]. In general, urban households were more likely to own these assets than rural households. Improved hygiene and sanitation, which are indicative of SES and improved infrastructure [4,15,23] have been associated with better nutritional outcomes in previous studies [23], a finding corroborated by our study.

Food crises in SSA likely contribute to high prevalence rates of underweight, which was on the decline in Tanzania but is now increasing [25]. This raises public health concerns given that Tanzania has one of the largest rural-urban disparities in malnutrition [17], fueled by a falling rate of urban malnutrition despite rapid growth in cities [17] and limited progress on rural development. From 2004 to 2016, there was a slight increase in rural Tanzanian households with electricity (1% vs. 5%), sanitation facilities (1% vs. 14%), and access to improved water sources (22% vs. 25%) [24,26]. Rural and urban health is intricately linked, and impacted by accessibility to health care, improvements in infrastructure, and development policies.

This study responds to WHO and UNICEF requests for more research on younger adolescents, who are at a vulnerable stage in the life course due to sexual maturation, emerging health risks, and integration into adult lifestyles. Using a rigorous analytical framework, this study provides further evidence of an urban advantage in BAZ and HAZ among young adolescents in SSA, after controlling for household assets, household living conditions, and other covariates. Corroborated by previous studies [2,5,8,12], our results found that stunting was more pronounced in rural and urban males than in females.

The study design and large sample size are important strengths that add to the robustness and generalizability of this study’s findings. The main limitation of this study is its cross-sectional design, which limits causal inference. Other limitations result from merging data from two separate studies in different regions of the country. The urban study utilized multilevel modeling to account for variation due to clustering of participants. Hence, the representativeness of samples varies due to different spatial diffusion of participants between the two studies. However, since adolescents in both studies were randomly selected, merging these datasets for comparative analysis was scientifically grounded and in alignment with previous epidemiological approaches.

Similarly, although the present study data was collected from two different regions in Tanzania, the disaggregated data on each district is informative and the comparison of districts is vital in providing country-level assessment. Furthermore, the comparison of nutritional status between Moshi and Kilosa seems appropriate based on the similar mean z-scores (±SD) for weight-for-height, height-for-age, and weight-for-age among children 0–59 months of age [27] in both Kilimanjaro (where Moshi is situated) and Morogoro (where Kilosa is situated) regions [28]. In lieu of a national reference for nutrition and health parameters, studies have recommended the use of health data from the Kilimanjaro region (and Moshi urban municipality) as a proxy for a national reference [29,30]. Using this framework, and by comparing the adolescents to the internationally-recognized WHO growth reference, we found that the actual difference in nutritional status between the two groups was striking and justifies our rural/urban comparison.

Another limitation of comparing data from two different regions may include the wide range of ethnic communities involved; Tanzania represents the highest level of ethnic diversity in SSA [31,32] with more than 120 ethnic or tribal groups. However, it is important to note that most of the population, including those in Kilosa and Moshi districts, is of Bantu heritage [33]. A high level of ethnic diversity and some overlap in tribal groups was reflected in both Kilosa and Moshi districts in the most recent national data that includes tribal or ethnic identity [34].

Maternal education, income, and food security status indicators, which have been associated with improved child nutritional outcomes and partly explain the urban health advantage [2,4,5,17], were unavailable in this study. These variables may have provided additional insight on BAZ and HAZ differentials observed across rural/urban settings. Subjective self-reports of health status and sociodemographic indicators may result in upward bias in the estimated effect of these variables on undernutrition and stunting. Bias was minimized through rigorous processes in survey item development, question order, interviewer training, piloting and revising the survey, as well as querying both adolescents and their guardians. However, we were not able to fully utilize self-reported health data considering assessment methods differed across the datasets with the rural study employing a WHO clinical reference on self-reported health illnesses and the urban study assessing susceptibility to illnesses. Finally, assets varied with rural data reflecting a broader range of assets salient to this setting and urban data reflecting material assets. Since assets common to both geographic settings favored material assets, an urban bias may be reflected in this measurement.

Conclusions

Undernutrition and stunting are global concerns that impact long-term health, quality of life, and productivity. The prevalence of undernutrition and stunting observed in this study, particularly in rural areas, raises concerns for adolescent health. The life course theoretical model (LCT), widely accepted in maternal and child health, posits that cumulative effects of biological, social, economic, and environmental risk factors are the underlying causes of persistent health inequities and poor health outcomes. However, conventional application of LCT is focused on the perinatal period and fails to connect the critical life stage of adolescence, an important stage of developmental processes that links childhood and young adulthood [35]. Nutrition research generally focuses on individual level factors and a comprehensive integration of social determinants of health, such as the place of residence, is needed to elucidate factors that may be addressed beyond the individual level [36].

Long-term improvements in the social determinants of malnutrition, such as SES, education, and rural infrastructure [2-5], can sustain gains in population health. Although child malnutrition is more likely to occur in poorer households in SSA [2,4,13-15,17,22], some analyses across SES levels have determined that intra-urban nutritional disparities are greater than intra-rural disparities [14,15]. Better understanding of social determinants of health, including identifying factors specific to place of residence and the extent to which these factors impact adolescent health, and situating these findings in the context of rural to urban migration, would be instrumental for developing policies to close the gap on rural-urban disparities.

Study findings indicate that both place of residence and an individual’s status within that place are important determinants of nutritional status. Health interventions targeting rural populations, especially those in the lowest SES quintile, may have the greatest impact on improving the health of vulnerable adolescents. Finally, stunting-related obesity in urban areas should be carefully examined given Tanzania’s high urban growth rate [10,17,37], high prevalence of rural stunting [11,15,21], and migration patterns [10,37]. National policies need to pay specific attention to the dual burden of disease stemming from the occurrence of undernutrition and overweight, and develop effective strategies for the prevention of chronic diseases [2]. Early adolescence provides critical opportunities to employ these strategies for health promotion and disease prevention.

Moshi caregiver survey.

(DOC)

Click here for additional data file.

Moshi adolescent survey.

(DOC)

Click here for additional data file.

Kilosa adolescent survey.

(DOC)

Click here for additional data file.

Kilosa caregiver survey.

(DOC)

Click here for additional data file.

16 Aug 2021

PONE-D-21-20245

Rural-urban disparities in the nutritional status of younger adolescents in Tanzania

PLOS ONE

Dear Dr. Cordeiro,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

ACADEMIC EDITOR: Considering my own reading of the paper and reviewers opinion, I am in favour of recommending this paper subject to the revisions as suggested by reviewer-2.

==============================

Please submit your revised manuscript by Sep 30 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Srinivas Goli, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please amend your current ethics statement to address the following concerns:

a) Did participants provide their written or verbal informed consent to participate in this study?

b) If consent was verbal, please explain i) why written consent was not obtained, ii) how you documented participant consent, and iii) whether the ethics committees/IRB approved this consent procedure.

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. If the original language is written in non-Latin characters, for example Amharic, Chinese, or Korean, please use a file format that ensures these characters are visible.

4. Please state whether you validated the questionnaire prior to testing on study participants. Please provide details regarding the validation group within the methods section.

5. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

6. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

Considering my own reading of the paper and reviewers opinion, I am in favour of recommending this paper subject to the revisions as suggested by reviewer-2.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This study has touched an important and interesting topic on adolescents. But why it has considered the age group of 10-14 years only, while, the WHO reports mentioned 10-19 years. Is there any particular reason to take this age group only?

Reviewer #2: 1. While the researcher argue that there is lack of information on urba-rural adolescents nutrition, Their literature review shows that studies of that type were conducted in Sub-saharan Africa. This includes studies in Nigeria, Cameroon and Ethiopia. This create contradiction which calls for more thorough review to ensure that the gap is clearly seen.

2. I feel that this study has applied very inapropriate comparison. It could have been importat to compare urban rural within one region at least to capture advantages of one area against another area sharing similar characteristics but diffferent in geographical setting. Kilosa rural is too far from Moshi and has people of different cultures and other developmental milestones Basing on this grount, the differences between Kilosa and Moshi are too obvious and does not warrrant any study. It could have been prroper if Moshi urban was compared against Moshi rural where people are similar but has differences in Geography. Or Comparison could have been executed between Kilosa rural and Kilosa urban to see the differences. Further another way could have been comparing Kilosa rural and Morogoro which is the capital of the same region.

3. Kilosa and Moshi are in areas of too different diesease profile. Whereas Kilosa is in area of high prevance of communicable disease, Moshi is not. For instance,prevalance of Malaria in Moshi is very low as compared to high prevance in Kilosa. This has serious implication in nutrition and has been documented in other national reports. It could be good if the comparison was conducted in areas where there is commonalities. the health parameters used for comparison are innapropriate and cannot provide credible results

4. School enrollment and other developmental milestones between Kilosa and even rural areas of Moshi are too different from each other. One should expect more differences between Kilosa rural and Moshi urban. I feel that some of the parameters used for comparison have yielded very obvios results

5.The differences between Kilosa and Moshi are too obvious and therefore, if one is to make comparison, tha analysis should focus on parameters which are not obvious. Otherwise, the comparison should be within same zone if the need is to compare geographical areas. It is crucial for researchers to think on how best to present this comparison avoiding geographical comparisons in areas which are not sharing common characteristics

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

9 Nov 2021

Also included in word format under 'attach files'.

September 27, 2021

Dear Dr. Goli,

We found the comments of the reviewers to be thorough and helpful in improving our manuscript. We have carefully considered and addressed each comment as outlined below. We thank you and the reviewers for the comments which have significantly improved the manuscript.

We have also included out Financial Disclosure statement below:

This study was funded with the support of UNICEF/ Tanzania; U.S. National Institute of Mental Health (R01 MH66801), and the University of Massachusetts Amherst. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Best,

Lorraine S. Cordeiro and Nicholas P. Otis

Response to Editorial Comments

1. Formatting has been addressed.

2. Ethics: Inserted statement about written informed assent and consent in the Ethics Section of submission and in manuscript.

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. If the original language is written in non-Latin characters, for example Amharic, Chinese, or Korean, please use a file format that ensures these characters are visible.

Inserted: Adolescents and their caregivers were interviewed at the adolescent's residence or schools. Surveys in both English and Swahili are included as supporting information.

Supporting Information: S1 File: Kilosa survey in English and S2 File: Moshi survey in English.

Please state whether you validated the questionnaire prior to testing on study participants. Please provide details regarding the validation group within the methods section.

Inserted: The questionnaires were pre-tested among a sample of non-participants in Kilosa District and Moshi prior to using it with study participants. Trained interviewers administered the structured, pre-tested questionnaires in Swahili. The survey tool was developed in English based on an extensive literature review, and translated into Swahili by language experts in Tanzania, and then back-translated into English. We conducted content validity, ensuring that terms and meanings were understood by participants in the pretest and that terms were accurately translated from English to Swahili. Construct validity required setting a priori hypotheses of associations and analyzing the pilot data to see if these hypotheses signaled validity. For quality assurance, piloting the survey tool among adolescents and their parents/guardian in both Kilosa and Moshi resulted in no major changes in the content and construct of the survey.

4. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match. When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

Authors Response: Corrected funding information and financial disclosure so they matched.

This statement is required for submission and will appear in the published article if the submission is accepted. Please make sure it is accurate and that any funding sources listed in your Funding Information later in the submission form are also declared in your Financial Disclosure statement.

Authors Response: The authors have declared that no competing interests exist.

5. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Describe where the data may be found in full sentences. If you are copying our sample text, replace any instances of XXX with the appropriate details.

Authors Response: All urban data files and research materials are available from the Harvard University Data Repository at https://dataverse.harvard.edu/. Access to rural data can be requested directly from the corresponding author and surveys are included in this manuscript under Supporting Information.

6. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Authors Response: The reference list is complete and correct.

Response to Reviewers

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Authors response: No response is needed for this question.

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Authors response: No response is needed for this question.

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Authors response: All urban data files and research materials are available from the Harvard University Data Repository at https://dataverse.harvard.edu/. Access to rural data can be requested directly from the corresponding author and surveys are included in this manuscript under Supporting Information.

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

Authors response: No response is needed for this question.

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters).

Reviewer #1: This study has touched an important and interesting topic on adolescents. But why it has considered the age group of 10-14 years only, while, the WHO reports mentioned 10-19 years. Is there any particular reason to take this age group only?

Authors response: We appreciate this comment from Reviewer #1. The urban dataset only included adolescents aged 10-14 years, while the rural dataset included individuals aged 10-19 years. For comparative purposes, this study examined data on adolescents aged 10-14 years which was available in both datasets. We justify this focus on younger adolescents based on the WHO recommendation for more research on this age group (i.e. <15 years of age). WHO states that “there is a pressing need for research in this area in low and middle income countries to understand about the realities of young adolescents lives, especially in the context of the rapid changes that are occurring in societies.”1 Furthermore, the Guttmacher Institute identified key research gaps in health information noting that “excluded groups of adolescents in developing regions include adolescents younger than 15, unmarried/never-married women, youth in vulnerable situations, male adolescents...”2 Data on 10-19 year old participants in the rural dataset are published elsewhere.3

1. https://www.who.int/reproductivehealth/topics/adolescence/very_young_ados/en/

2. https://www.guttmacher.org/report/research-gaps-in-sexual-and-reproductive-health

3. Cordeiro LS, Wilde PE, Semu H, Levinson FJ. Household food security Is inversely associated with undernutrition among adolescents from Kilosa, Tanzania. Journal of Nutrition, 2012;142: 1741–1747. doi:10.3945/jn.111.155994

Reviewer #2:

1. While the researcher argue that there is lack of information on urba-rural adolescents nutrition, Their literature review shows that studies of that type were conducted in Sub-saharan Africa. This includes studies in Nigeria, Cameroon and Ethiopia. This create contradiction which calls for more thorough review to ensure that the gap is clearly seen.

Authors response: Reviewer#2 raised a valid and important issue. We conducted a comprehensive review of the literature based on the comments from Reviewer #2. We concluded that a robust literature base is still in its infancy in regards to adolescent nutrition in Sub Saharan Africa (and truly adolescents in general); however, literature on this population subgroup is steadily growing. There are relatively few adolescent studies on urban-rural comparisons in this region of the world, and most are either in West Africa or in Ethiopia. Two studies in Ethiopia, one in Nigeria, and one in Cameroon are presented in this manuscript, indicating limited studies available to draw conclusions about urban-rural disparities in adolescent nutritional status across Sub Saharan Africa. This study presents new data from Tanzania, a low-income East African country with a political economy and demographical profile that is different from Ethiopia and West African nations, and adds to the literature base on adolescent nutrition and urban-rural disparities in Sub Saharan Africa.

2. I feel that this study has applied very inapropriate comparison. It could have been importat to compare urban rural within one region at least to capture advantages of one area against another area sharing similar characteristics but diffferent in geographical setting. Kilosa rural is too far from Moshi and has people of different cultures and other developmental milestones Basing on this grount, the differences between Kilosa and Moshi are too obvious and does not warrrant any study. It could have been prroper if Moshi urban was compared against Moshi rural where people are similar but has differences in Geography. Or Comparison could have been executed between Kilosa rural and Kilosa urban to see the dfferences. Further another way could have been comparing Kilosa rural and Morogoro which is the capital of the same region.

Authors response: We agree with Reviewer # 2 that comparing adolescents in the same region of the country would have been optimal. Reviewer #2 raises an issue that allows us to strengthen our paper and we appreciate the thoughtful comments. To address the concerns raised by Reviewer #2, we present our rationale for the comparison between Kilosa and Moshi below.

a. Tanzania represents the highest level of ethnic diversity in SSA1,2 with 120+ ethnic or tribal groups, however, it important to note that most of the population is of Bantu heritage.3 So, while Kilosa District is in a different region than Moshi, it comprises a large population of the dominant ethnic groups found in Moshi District and a vast majority of the population in both districts is of Bantu heritage. There were 54 ethnicities represented in the Kilosa rural dataset and the Moshi urban data also had a high level of ethnic diversity. We also decided to examine national data more closely and requested access to DHS data. In an analysis of the 1996 DHS dataset representing the most recent national DHS survey that includes tribal or ethnic identity, there were over 25 different tribes in both Kilimanjaro and Morogoro regions with overlap in several groups across the regions. For context, Miguel (2004) notes that Tanzanian leadership focused on national unity across ethnic identities soon after independence4, leading to insignificant effects in political representation and financial distribution that could be caused by politicization of ethnicity.1 We also recognize that there is a lack of country data disaggregating health outcomes by ethnicity due to the Tanzanian socialist ideology that has successfully emphasized national identity over ethnicity.5

1.Weber, A. (2010). The causes of politicization of ethnicity: A comparative case study of Kenya and Tanzania. APSA 2010 Annual Meeting Paper. doi:10.5167/uzh-63126

2.Fearon, J. D. (2003) Ethnic and Cultural Diversity by Country. Economic Growth, 8, 195–222. doi:10.1023/A:1024419522867

3.Goldberg 2020 Country-Report-Tanzania_Goldberg.pdf. (n.d.). Retrieved August 8, 2021, from https://rad-aid.org/wp-content/uploads/Country-Report-Tanzania_Goldberg.pdf

4.Miguel, E. (2004). Tribe or nation? Nation building and public goods in Kenya versus Tanzania. World Politics, 56(3), 327-362. doi:10.1017/S0043887100004330

5.Lawson, D. W., Borgerhoff Mulder, M., Ghiselli, M. E., Ngadaya, E., Ngowi, B., Mfinanga, S. G., ... & James, S. (2014). Ethnicity and child health in northern Tanzania: Maasai pastoralists are disadvantaged compared to neighbouring ethnic groups. PloS one, 9(10), e110447. doi:10.1371/journal.pone.0110447

b. The comparison of the rural and urban data presents one focus on this manuscript and the main focus is the findings from each area of Tanzania. In lieu of a national reference for nutrition and health parameters, studies have presented health data from the Kilimanjaro region (and Moshi Urban Municipality) and recommend the use of this data as a proxy for a national reference. For example, in a study on infants, children, and adolescents in Moshi, Kilimanjaro region, Buchanan et al. (2010) reported that “data regarding immunological and haematological reference intervals for healthy African populations are scarce, particularly for infants, children, and adolescents. Values currently used are often based upon results generated from Caucasian populations living in industrialized countries (Wintrobe 1981; Tugume et al. 1995; Karita et al. 2009).”1 Their study aimed to “either verify or establish normal reference ranges, for haematological and immunological indices among healthy Tanzanian children…[and their study provided] further evidence that the establishment of local reference ranges is critical for optimal patient management and medical research. Reference intervals derived primarily from Caucasians residing in developed nations, in particular, are inappropriate for this population.”1 They also provide biochemistry reference values using data for healthy children and adolescents in the Kilimanjaro region of Tanzania.2 Using this framework and by comparing the Tanzanian rural and urban adolescents to the internationally-recognized WHO growth reference, we note that the actual difference in nutritional status between the two groups is striking and justifies that this comparison presents valuable information on disparities in Tanzania.

1.Buchanan, A. M., Muro, F. J., Gratz, J., Crump, J. A., Musyoka, A. M., Sichangi, M. W., ... & Cunningham, C. K. (2010). Establishment of haematological and immunological reference values for healthy Tanzanian children in Kilimanjaro Region. Tropical Medicine & International Health, 15(9), 1011-1021. doi:10.1111/j.1365-3156.2010.02585.x

2.Buchanan, A. M., Fiorillo, S. P., Omondi, M. W., Cunningham, C. K., & Crump, J. A. (2015). Establishment of biochemistry reference values for healthy Tanzanian infants, children and adolescents in Kilimanjaro Region. Tropical Medicine & International Health, 20(11), 1569-1577. doi:10.1111/tmi.12580

3. Kilosa and Moshi are in areas of too different diesease profile. Whereas Kilosa is in area of high prevance of communicable disease, Moshi is not. For instance, prevalance of Malaria in Moshi is very low as compared to high prevance in Kilosa. This has serious implication in nutrition and has been documented in other national reports. It could be good if the comparison was conducted in areas where there is commonalities. the health parameters used for comparison are innapropriate and cannot provide credible results.

Authors response:

a. We addressed the concerns regarding comparison of the two districts above.

b. Since self-reported health is generally a poor index of health and not necessarily a salient predictor of undernutrition in this area, we recommend that the health parameters based on self-reported data be viewed in this context and have discussed this further in our limitations.

c. In 2018, 94% of global malaria deaths occurred in SSA1. In 2018, case fatality of malaria contributed to 18% of under five mortality in SSA1 and malaria likely has a negligible effect on undernutrition in adolescence. Research on the malaria-malnutrition interactions indicate a complex relationship with inconclusive findings. Some studies report no association between malaria and malnutrition2 while others find an association. Das et al. (2015) concluded from a systematic review of the literature that “the evidence on the effect of malnutrition on malaria risk remains inconclusive… Further clarification on malaria-malnutrition interactions would also serve as a basis for designing [interventions].”3

1.Ouédraogo, M., Kangoye, D. T., Samadoulougou, S., Rouamba, T., Donnen, P., & Kirakoya-Samadoulougou, F. (2020). Malaria case fatality rate among children under five in Burkina Faso: an assessment of the spatiotemporal trends following the implementation of control programs. International journal of environmental research and public health, 17(6), 1840. doi:10.3390/ijerph17061840

2.Charchuk, R., Houston, S., & Hawkes, M. T. (2015). Elevated prevalence of malnutrition and malaria among school-aged children and adolescents in war-ravaged South Sudan. Pathogens and global health, 109(8), 395-400. doi:10.1080/20477724.2015.1126033

3.Das, D., Grais, R. F., Okiro, E. A., Stepniewska, K., Mansoor, R., Van Der Kam, S., ... & Guerin, P. J. (2018). Complex interactions between malaria and malnutrition: a systematic literature review. BMC medicine, 16(1), 1-14. doi:10.1186/s12916-018-1177-5

d. We note that the comparison nutritional status between Moshi and Kilosa seems appropriate based on the similar mean z-scores (±SD) for weight-for-height, height-for-age, and weight-for-age among children 0-59 months of age1 in both Kilimanjaro (where Moshi is situated) and Morogoro (where Kilosa is situated) regions.2

Kilimanjaro Morogoro

Weight for height 0.05 ± 0.99 -0.06 ± 1.09

Height for age -1.09 ± 1.12 -1.26 ±1.13

Weight for age -0.58 ± 1.05 -0.74 ± 1.09

Source: Tanzania National Nutrition Survey 20182

1.https://www.who.int/tools/growth-reference-data-for-5to19-years 2.https://www.unicef.org/tanzania/reports/tanzania-national-nutrition-survey-2018

4. School enrollment and other developmental milestones between Kilosa and even rural areas of Moshi are too different from each other. One should expect more differences between Kilosa rural and Moshi urban. I feel that some of the parameters used for comparison have yielded very obvios results.

Authors response: We appreciate this relevant comment from Reviewer #2 and have presented the rationale for the comparison between Kilosa and Moshi in our response to comments #2 and #3 above.

5.The differences between Kilosa and Moshi are too obvious and therefore, if one is to make comparison, tha analysis should focus on parameters which are not obvious. Otherwise, the comparison should be within same zone if the need is to compare geographical areas. It is crucial for researchers to think on how best to present this comparison avoiding geographical comparisons in areas which are not sharing common characteristics.

Authors response: We appreciate this relevant comment from Reviewer #2 and have presented the rationale for the comparison between Kilosa and Moshi in our response to comments #2 and #3 above.

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Authors response: No response is needed for this question.

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Authors response: All figures were uploaded to PACE and included in this revision.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

3 Dec 2021

Rural-urban disparities in the nutritional status of younger adolescents in Tanzania

PONE-D-21-20245R1

Dear Dr. Cordeiro,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Srinivas Goli, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Considering the reviewers suggestion and my own reading of this paper, I am recommending it for publication in PLOS One.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: The authors have adressed most of the comments in a satisfactory manner . This paper may be ready for publication

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

9 Dec 2021

PONE-D-21-20245R1

Rural-urban disparities in the nutritional status of younger adolescents in Tanzania

Dear Dr. Cordeiro:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Srinivas Goli

Academic Editor

PLOS ONE