Infection, nutritional status, and body composition: Associations at birth and 6 months postnatally in Soweto, South Africa

The impact of infection on infant nutritional status, body size, and growth is well documented. However, research into the impact of infection on infant body composition is limited. Greater understanding is, therefore, needed on the effects of infection in early life.

in infants in the first 6 months postnatally, and that these efforts should concentrate on access to safely managed sanitation facilities.

| INTRODUCTION
Child undernutrition is a major global health concern with an estimated 144 million children under the age of 5 being stunted, and 47 million wasted (UNICEF et al., 2020). It is estimated that, worldwide, around 249 million children under 5 years of age are at risk of not reaching their developmental potential (Lu et al., 2016).
The first 1000 days (from conception to 2 years of age) is an integral phase of childhood development due to developmental plasticity, making it the most pertinent period for interventions aimed at optimizing growth and development (Adair et al., 2013;Norris et al., 2017;Prendergast et al., 2014;Victora et al., 2010). Developmental plasticity posits that environmental conditions experienced in early life, including nutrition, trigger permanent physiological adjustments that can profoundly influence human biology and long-term health outcomes (Hochberg et al., 2011;Kuzawa, 2005;Said-Mohamed et al., 2018).
Stunting among children in Africa has decreased in prevalence from 38. 3% (2000) to 30.3% (2017), however, because of population growth, the actual number of stunted children over the same period has risen (Independent Expert Group of the Global Nutrition Report, 2018). At the same time, in percentage terms, stunting has declined twice as quickly in Asia, Latin America, and the Caribbean as it has in Africa (UNICEF et al., 2017). Africa is, therefore, the only region globally that has witnessed an increase in stunting prevalence in children under 5 years of age, and strikingly, the probability of a child being underweight has increased by almost 50% since 2000 (FAO et al., 2018;UNICEF et al., 2020).
In contrast to other countries in sub-Saharan Africa, South Africa has a long history of recording nutrition data through studies and national surveys and is also in an advanced stage of nutrition transition associated with various political, economic, and demographic transitions (Said-Mohamed et al., 2015). Important in this regard is the substantial level at which undernutrition persists in South African children under 5 years of age, where the prevalence of stunting has been reported at 26.5%, underweight at 6.1%, and wasting at 2.2% (Said-Mohamed et al., 2015;Shisana et al., 2013).
Infection and undernutrition are closely interlinked, with undernutrition proposed as a primary cause of immunodeficiency, increasing the likelihood of an infant contracting an infection (Katona & Katona-Apte, 2008). In addition, frequent illnesses in infancy can impair nutritional status, further predisposing the infant to more infections (Dewey & Mayers, 2011).
Infections may decrease food intake, impair nutrient absorption, cause direct nutrient losses, increase metabolic requirements or catabolic losses of nutrients and, possibly, impair transport of nutrients to target tissues.
When considering infection, one cannot avoid mention of inflammation which has been demonstrated to be a critical component of innate immune defenses against infection and injury. This is characterized by the rapid recruitment of immune cells to sites of infection and inflammation through the production of cytokines and chemokines (Marshall et al., 2018). Acute activation of inflammatory processes after pathogen exposure is rapid, usually within hours, whereas more specific adaptive immune processes take several days (McDade, 2012).
In addition, induction of the acute phase response and production of proinflammatory cytokines may directly affect growth (Stephensen, 1999). Growth and endochondral ossification have for instance been found to be inhibited by pro-inflammatory cytokines including the activin A-follistatin system, glucocorticoids, and fibroblast growth factor 21 (FGF21), all of which are elevated during infection (Millward, 2017).
Immune function, including mounting responses to infection, is energetically costly and potentially diverts calories away from less immediately essential functions. It has, therefore, been proposed that among children, the allocation of energy toward immune function may lead to trade-offs with physical growth, particularly in highpathogen, low-resource environments (Urlacher et al., 2018). Different types of infections have been shown to have different energetic costs, with those resulting from soil transmitted helminths having the greatest effect on growth (Gildner et al., 2022;Urlacher et al., 2018). In addition, undernutrition has been shown to have immunosuppressive effects which in turn have been argued to, in certain instances, have an adaptive effect in so far as downregulation of immune activity may free up resources for other biological functions, including growth (McDade, 2005;Urlacher et al., 2018). In contrast high pathogen environments postnatally are associated with upregulated immune activity which is energetically costly and likely to impact on growth (McDade, 2005). Multilevel analyses have demonstrated consistent negative effects of immune activation on growth. Studies have for instance shown how children with greater levels of body fat (i.e., energy reserves) were able to mitigate the growth-inhibiting effects of acute inflammation (Urlacher et al., 2018). The role that body composition, and body fat in particular plays, therefore, emerges as a significant mediating factor.
With regards to fat mass (FM), this has been linked to the provision of metabolic precursors related to energy and the secretion of leptin, for immune function (Lord, 2002;J. Wells, 2010). Therefore, in cases of undernutrition, a reduction in FM is often observed which could be the case in infection too. In addition, in cases of severe-acute undernutrition, fat distribution was noted to be more central, mainly due to the loss of peripheral fat (J. C. K. Wells, 2019).
It has been predicted that in cases of undernutrition, a decrease in fat-free mass (FFM) could represent mobilization of protein from muscle mass when characterized by a lack of dietary protein, to provide crucial proteins required for immune function (J. C. K. Wells, 2019). Furthermore, studies suggest that organs are protected at the expense of muscle (Prentice, 1999). More chronic cases of undernutrition have been associated with deficits in FFM, especially relating to aspects like organ size (Ece et al., 2007;J. C. K. Wells, 2019). Severe wasting has been shown to affect fat and fat-free tissues and that while levels of fat often recover, fat-free tissues may remain lower in the long term (J. C. K. Wells, 2019).
Top causes of infant mortality and morbidity in South Africa include diarrhoeal disease, meningitis, lower respiratory tract infections such as pneumonia, perinatal conditions associated with HIV and AIDS, and malnutrition (Chola et al., 2015;Mabaso et al., 2014). Diarrhea is one of the leading causes of morbidity and mortality in under-five children in South Africa, though the true burden of childhood diarrhea is not known (Chola et al., 2015). The 2010 General Household Survey (GHS) showed that there were over 60 000 cases of childhood diarrhea per month (Chola et al., 2015;Statistics South Africa, 2012). The GHS is conducted every year following a multi-stage, stratified random sampling, with approximately 30 000 households visited twice every year (Statistics South Africa, 2012). Another leading cause of morbidity and mortality among children under 5 years of age in South Africa is respiratory diseases, where in the same survey, almost 12% of children were reported to have had respiratory tract infections (Statistics South Africa, 2012).
Diarrhea and poor sanitation during the first 2 years of life have been associated with elevated C-reactive protein (CRP) in young adulthood, independent of linear growth (Said-Mohamed et al., 2019). As a result, it has been suggested that early-life exposure to infections can predict greater levels of inflammation in adulthood .
Body composition in infancy has been found to influence future body composition in adulthood (Admassu et al., 2018). For instance, weight gain in the first 1000 days is associated with increased adult FFM, whereas weight gain in later childhood has been linked to an increase in adult fat mass (Victora et al., 2008). Catch-up growth (rapid growth during infancy in response to a period of growth inhibition) has also been associated with increased obesity risk in childhood and later life (Ong, 2007). Increased adult fat mass increases the risk of overweight, obesity, and noncommunicable diseases in adulthood. In addition, catch-up growth has been linked to increased insulin resistance, which has also been associated with changes in body composition and growth (Ong, 2007).
Given the array of effects and interactions between infection, inflammation, body composition, nutritional status, and their respective long-term consequences, it is important to explore the relationship between infection (frequency and intensity) and nutritional status and body composition during early life. Within the age range of birth to 6 months, there are limited studies (Soares et al., 2019) on the effect of infection on body composition and no studies in a sub-Saharan African population. Commonly weight-for-age (WAZ), weight-for-height (WHZ), and/or body mass index (BMI) are used to assess the impact of infection on the body; however, these metrics are unable to identify which component of body composition is affected. This paper, therefore, intends to shed light on the impact of infection, specifically on infants' body composition (FM, FFM, FMI and FFMI) and nutritional status (HAZ and WHZ) between birth and 6 months postnatally.

| Study design
This study comprises analysis of data from a prospective cohort study consisting of mother-infant pairs who were followed up weekly from birth to 6 months. This study comprises a cross-sectional analysis of anthropometric and body composition measures at 6 months postnatally. This data originates from a larger study entitled "Interaction between nutrition, infection, household environment and care practices and their impact on growth and development in infants between birth and one year of age," colloquially also known as the Soweto Baby WASH (SBW) study (Momberg et al., 2020). The aims of the SBW study were to record maternal and infant morbidity and illness, household environment, and infant feeding and care practices as well as assess associations with infant nutritional status and development between birth and 1 year postnatally. The SBW study included data pertaining to infant and maternal anthropometric and body composition measurements, household water, sanitation, and hygiene (WASH), maternal and infant morbidity, illness and healthcare access, infant feeding practices, household socio-economic status and demographics.
Recruitment began in January 2018, and the last point of data collection for this study was in September 2018, when the last recruited infant turned 6 months of age. The screening and recruitment took place at the maternity services of the Chris Hani Baragwanath Academic Hospital (CHBAH) in Soweto, Johannesburg, in South Africa.

| Study setting
Soweto, situated in the City of Johannesburg in the Gauteng Province, is the largest township in South Africa with an estimated population of 1.3 million residents, with between 600 000 and one million people regarded as living in extreme poverty (Government of the Republic of South Africa, 2020; Harrison & Harrison, 2014). CHBAH is the only tertiary academic teaching hospital servicing this community and is the third largest hospital in the world with approximately 3200 beds. The maternity facilities at CHBAH service approximately 60 000 patients per year (Department of Health, 2023). In 2018, it was reported that Gauteng had near universal water and sanitation coverage with 97.7% of the population having access to piped tap water in their dwellings, and 91.8% having access to improved sanitation, defined as those that hygienically separate human waste from human contact (Government of the Republic of South Africa, 2020; Moeti & Padarath, 2019).

| Ethical considerations
The Soweto Baby WASH Study received clearance from the University of the Witwatersrand's Human Research Ethics Committee (Medical) (Certificates: M170753, M170872, and M170955) (Momberg et al., 2020). In addition to this, written informed consent was obtained from all participants.

| Participants
The inclusion criteria for mother and infant pairs in the Soweto Baby WASH study were mothers ≥18 years of age at the time of screening, singleton pregnancy, infant birthweight between ≥2500 g and <4000 g and a term pregnancy between ≥38 weeks and <42 weeks, and willing and able to give informed consent for participation in the study. Infants that had been diagnosed with physical, mental, or congenital disorders at birth and mothers living with HIV/AIDS were not included in the study. A total of 1289 mother-infant pairs were screened postdelivery at the maternity wards at CHBAH. Of those screened, 243 were eligible and 156 mother-infant pairs consented and were enrolled in the study.

Infant morbidity and illness
To collect data regarding infant morbidity and illness, an interviewer-administered questionnaire was used and confirmed against clinical records. This questionnaire was adapted to match the study setting. Due to the weekly visits up to 6 months, any illness episodes within a 7-day period were recalled and recorded for that week (Momberg et al., 2020).
The following variables were used to represent infection and morbidity in the infants: fever, irritability, coughing/wheezing/hoarseness, difficulty breathing, sneeze/runny/stuffy nose, diarrhea, vomiting, and rash. These variables were then used to construct a composite morbidity index consisting of the sum of the cumulative tallies for each symptom, divided by the number of visits they had, as an increased number of visits is more likely to detect more symptoms by nature.

| Covariates
Maternal characteristics: Maternal anthropometry Using a wall mounted Holtain stadiometer, maternal height was measured to the nearest 0.1 cm and, using a Seca 877 digital scale, weight was measured to the nearest 0.1 kg (Momberg et al., 2020). This was done by trained research assistants using standardized techniques (de Onis et al., 2014). These included the calculation of Body Mass Index (BMI) as (kg/m 2 ) (WHO Child Growth Standards, 2009).

Infant characteristics: Infant feeding practices
An interviewer-administered questionnaire locally adapted from the World Health Organization/United Nations Children's Fund (WHO/UNICEF) Infant and Young Child Feeding Questionnaire was used to assess infant feeding practices (World Health Organization (WHO), 2008). The WHO definition for exclusive breastfeeding (EBF) was used to determine infant feeding practices up to 6 months (Momberg et al., 2020). The WHO definition allows breastmilk, expressed or from a wet nurse, oral rehydration solution, drops, and syrups to be classified as EBF, but nothing else (World Health Organization, 2008). Unfortunately, exclusive breastfeeding had to be excluded due to the fact none of the infants were exclusively breastfed at 6 months of age.

Household factors: Socio-economic status and demographics
The questionnaire for household socio-economic status was built from the South African National Income Dynamics Study (University of Cape Town, 2016) and the Living Conditions Survey (Statistics South Africa, 2017). Household crowding (i.e., number of people cohabiting), maternal employment and education-related data were recorded using this questionnaire. From this, a Household Wealth Index (HWI) was calculated using latent household variables, which represent household-owned assets, including bicycle, motorcycle, motor vehicle, fridge, microwave, washing machine, landline telephone, camera, cell phone, television, DVD-player, paid television subscription, computer/laptop and internet access, and housing characteristics were included in the index. These variables represent home and land ownership and the main type of energy used by the household (Rutstein & Johnson, 2004;Vyas & Kumaranayake, 2006). Using these data, the HWI was built using principal component analysis (Vyas & Kumaranayake, 2006). After formulating the HWI, the sample was divided in two categories: high HWI and low HWI (Momberg et al., 2020).
Water, sanitation, and hygiene (WASH) Household WASH data were collected using an interviewer-administered questionnaire based on the WHO/UNICEF Core Questions on Drinking-Water and A water index was built summing variables relating to the water source, access, treatment, and supply, which was scored from 0 to 4, with 0 representing the lowest level of water infrastructure and behaviors and 4 being the highest (Momberg et al., 2020). Variables describing water sources were split into two categories: safely managed and not safely managed. Safely managed water sources were defined as per the JMP, as a basic drinking water source located on the premises (piped into the home/property), available when needed, and free of fecal and chemical contamination (WHO/UNICEF Joint Monitoring Programme (JMP), 2015). This is measured, monitored, and reported by the National Department of Water and Sanitation. Any other installations were described as not safely managed which include dug wells, surface water and the like. Other variables like access to a water source on the premises, type of water treatment employed at home, and frequency of interruptions to water supply were also collected and included in this index.
A sanitation index was built by placing sanitation infrastructure into the same two categories: safely managed and not safely managed. The safely managed sanitation facilities were defined as a basic sanitation facility (flush/pour to piped sewer system) where excreta are safely disposed of in situ or treated off-site (WHO/UNICEF Joint Monitoring Programme (JMP), 2015). Not safely managed sanitation facilities included all other infrastructure installations which refer to pit latrines, buckets, and open defecation.
A composite hygiene index was built, using a scoring system of 0-5, with 0 being the lowest level of hygiene behaviors and infrastructure and 5 being the highest (Momberg et al., 2020). To assess hygiene behaviors and circumstances, three indices were created (Webb et al., 2006): Household Hygiene Index was scored 0-2 and included the type of house (formal structure = 1 and informal structure = 0) and presence of animals on the property (yes = 1 and no = 0). The second index created was the Personal Hygiene Index, also scored 0-2, included handwashing (recommended handwashing practices = 1 and not recommended practices = 0) and handwashing detergent (with detergent = 1 and without detergent = 0). Finally, a Food Hygiene Index was created and scored 0-1, where the cleaning of the breast and cleaning of utensils prior to feeding was assessed (cleaned prior to feeding = 1 and not cleaned prior to feeding = 0). These three indices (Household Hygiene Index, Personal Hygiene Index and Food Hygiene Index) were calculated as the sum of the individual items and the overall Hygiene Index as the sum of the Household Hygiene Index, Personal Hygiene Index and Food Hygiene Index (Manzoni et al., 2019;Momberg et al., 2020;Webb et al., 2006).

| Outcome variables
Infant body size and nutritional status Trained research assistants used standardized techniques to obtain anthropometric measurements (Members of the WHO Multicentre Growth Reference Study Group, 2007). Infant weight was measured to the nearest 0.01 kg using a Seca 367 infant scale and recumbent length to the nearest 0.01 cm using a Seca 416 infantometer. Data were analyzed using the WHO Anthro Survey Analyzer software to generate height-for-age scores (HAZ), and weight-for-height (WHZ) scores (World Health Organization, 2020).

Infant body composition
The deuterium oxide dilution technique was used to assess infant body composition. Deuterium oxide ( 2 H 2 O) is water labeled with 2 H and is often denoted as D 2 O. The amount of deuterium in excess of natural abundance present before the dose was taken, is known as the enrichment of deuterium in body water. Total body water (TBW) is calculated from the enrichment of deuterium in the body water and the amount of deuterium oxide consumed.
Trained professionals orally administered 1 mL of deuterium oxide to the infants at 6 months of age. Saliva samples were collected as follows; 2 mL prior to deuterium dose (baseline), a 2 mL sample at 2:30 h post-dose, and a final 2 mL sample at 3:00 h post-dose. Fourier Transform Infrared Spectrometry (FTIR) was used to measure enrichment of deuterium from the saliva. The protocol for this was developed by the International Atomic and Energy Agency (IAEA) (International Atomic Energy Agency, 2009. The deuterium oxide dilution technique assessed infant body composition by providing an estimate of Total Body Water (TBW) (International Atomic and Energy Agency, 2010), and from this, the FFM was calculated using Fomon hydration coefficients (Fomon et al., 1982). Fat Mass was then calculated by finding the difference between body weight and FFM values, after which, fat free mass index (FFMI) and fat mass index (FMI) were calculated.

| Sample size
A total of 1289 mother-infant pairs were screened postdelivery at the maternity wards at CHBAH. Of those screened, 243 were eligible and 156 mother-infant pairs consented and were enrolled in the study. At 6 months postnatally, 64 mother-infant pairs had been lost to attrition, leaving a sample size of 92 mother-infant pairs ( Figure S1) (Momberg et al., 2020). Of the 92 that were still enrolled in study at 6 months, 82 were seen for their follow-up visit and 78 completed the deuterium dilution.

| Statistical methods
The statistical software used for analysis was Stata version 16.1 (StataCorp, 2019). Descriptive statistics, including frequencies and proportions were used to summarize categorical variables. A Skew test, students t-test, and Shapiro-Wilk test were used to determine whether continuous variables were normally distributed, following which, mean and standard deviation were used to summarize the variables, and where not, median and range were used to describe the data.
A social-ecological model (Neal and Neal 2013) informed by the UNICEF Conceptual Framework (The United Nations Children Fund, 1990) was used to construct three hierarchical models. These hierarchical models were hypothesized (Victora et al., 1997) to assess associations between morbidity and illness, and body composition outcomes, represented by FFM, FFMI, FM, and FMI, HAZ and WHZ, while controlling for infant and maternal factors, and household characteristics, including WASH components, from birth to the 6-month postnatally. In other words the hierarchical regressions seek to evaluate associations between the relative frequency of reported illness symptoms during the period between birth and 6 months and infant body composition at 6-month postnatally. Model 1 (M1) tested the unadjusted and independent association between morbidity and the aforementioned outcomes. Model 2 (M2) investigated this association after controlling M1 for individual infant factors (birthweight, gestational age, and sex). Model 3 (M3) examined the effect after controlling M2 for maternal factors (age, BMI, employment, and education). Lastly, Model 4 (M4) explored the effect after controlling M3 for household factors (water index, sanitation, hygiene index, and household wealth index). Collinearity between variables was tested by assessing the variable inflation factor and covariance correlation matrices. Significance levels were set at p ≤ 0.05 with 95% Confidence Intervals (CI).

| RESULTS
Infants without adverse birth outcomes were selected to better identify and investigate specific postnatal exposures and their impact on nutritional status and body composition. Birth outcomes, individual infant and morbidity characteristics, as well as maternal and household characteristics are reflected in (Table 1). Mean birthweight was 3.055 grams (SD = 0.35), and mean gestational age was 38.9 weeks (SD = 1.2). Attrition over the cumulative period from birth to 6 months postnatally accounted for a 41% reduction in the sample size described in Figure S1.
Over the study period, the most reported symptoms were sneezing/runny/snotty nose, indicative perhaps of upper respiratory tract infections (80.5%), followed by skin irritations/rash (70.5%), and coughing/wheezing/ Mean maternal BMI was 29.9 (SD = 6.1), and mean maternal age was 29.9 (SD = 6.6) at 6 months postnatally. More than two thirds (70.5%) of the mothers included in the study were employed at 6 months postnatally, while less than a quarter (22.8%) had some form of tertiary, post high school, qualification. Mean household size was 5.9 members (SD = 2.9), with near universal safely managed sanitation (96.3%), and 77.5% of households reporting a safely managed water source.
Tables 2-4 show the results for the hierarchical regression analyses on FFMI and FMI, FFM and FM, followed by HAZ and WHZ, at 6 months postnatally respectively. All significant associations for the various models are diagrammatically displayed in Figure 1.
Morbidity, over the cumulative period from birth to 6 months, was associated with lower FMI as well as lower FM in the unadjusted M1. For every 1-unit increase in the morbidity index, a decrease in FMI of À1.77 kg/m 2 (M1, p = 0.02) and a decrease in FM of À0.61 kg, (M1, p = 0.04) at 6 months was shown. Conversely, morbidity was also associated with an increase in FFM in the unadjusted M1 and adjusted M2. For every 1-unit increase in the morbidity index, an increase in FFM of 0.94 kg (M1, p = 0.001) and 0.49 kg, (M2, p = 0.01) at 6 months was shown.
No associations were found between the morbidity index and FFMI, HAZ, and WHZ.
Increased birthweight was associated with a higher FFM in M2 and M3, higher HAZ and WHZ in M2, M3, and M4. Finally, a safely managed sanitation facility was associated with a higher HAZ (β = 1.21, p = 0.02).

| DISCUSSION
This study aimed at assessing the relationship between morbidity and nutritional status and body composition in infants between birth and 6 months postnatally in Soweto, South Africa. In summary, morbidity, over the cumulative period of birth to 6 months, was associated with lower FMI (β = À1.77) as well as lower FM (β = À0.61), and conversely with higher FFM (β = 0.94), in infants at 6 months. No associations were found between the morbidity index and FFMI, HAZ, and WHZ. Increased birthweight was associated with a higher FFM (β = 0.66), HAZ (β = 1.14), and WHZ (β = 0.87). Finally, a safely managed sanitation facility was associated with a higher HAZ (β = 1.21). Overall, therefore, the results confirmed our hypothesis: that there is an association between infection and body composition in infants at 6 months of age. Evidence suggests that the greatest impact of morbidity between birth and 6 months on body composition is on FMI.
The detected decrease in FMI of À1.77 kg/m 2 and À0.061 kg in FM, per unit increase in the morbidity index has particular significance within this period of infancy. This is because the period of birth to 6 months is critical for brain growth and protecting body size (Ijaz & Rubino, 2012;Kuzawa, 1998;Kuzawa et al., 2014). Infection and the resultant relevant energy deficiency may trigger adaptive physiological mechanisms prioritizing brain growth at the expense of body size (Kubera et al., 2013;Kuzawa et al., 2014;Momberg et al., 2022;Urlacher et al., 2018). This may be mediated through adiposity and the provision of metabolic precursors, energy, and secretion of leptin for immune function and activation (Lord, 2002;Urlacher et al., 2018;J. Wells, 2010).
Recent commentary and review of existing literature reinforces the point that adiposity is more sensitive to fluctuations driven by external environmental factors, including food security, and recurrent infections, and therefore, more likely to affect children over the course of months rather than years (Momberg et al., 2022;Richard et al., 2013;Victora et al., 2010). Therefore, from a public health perspective, it is important to track variability in adiposity levels, throughout infancy and childhood, which are more responsive to infection and environmental fluctuations (Momberg et al., 2022). This may be why no association was detected with HAZ, which is rather a marker of chronic undernutrition.
Furthermore, a reduction in FMI, FM, in conjunction with exposure to inflammatory cytokines associated with mounting an immune response, could alter phenotypic trajectories of the infant due to this period of plasticity. This is because it is likely that adiposity is the source of energy that is mobilized in the trade-off of energy allocation, in turn having implications for long term public health risks, especially relating to noncommunicable diseases (Adair et al., 2013;Said-Mohamed et al., 2019).
The comparative increase in FFM in relation to morbidity somewhat complicates interpretation. In contrast to these findings, studies have predicted that in chronic cases of undernutrition, a decrease in FFM could represent mobilization of protein from muscle mass in cases characterized by a lack of dietary protein, to provide T A B L E 2 Summary of results for the hierarchical regression analyses of morbidity on FFMI and FMI at 6 months postnatally. crucial proteins required for immune function (Ece et al., 2007;J. C. K. Wells, 2019). However, it is our interpretation that in this instance, the increase in FFM is due to the automatic shift in proportion relative to the observed decrease in FM. The collinearity between the two parameters, therefore, limits the interpretation of absolute FFM and relative changes of FFM (Kyle et al., 2003). The use of WHZ is a commonly used marker of adiposity in infant body composition studies. We would, therefore, have expected to see associations with morbidity, the absence of which suggests that this measure may be less representative/sensitive to changes in adiposity levels in infants. The use of WHZ between birth and 6 months postnatally may be less likely to capture the effects of infection on the quantity of fat in infants. WHZ is also similar to BMI in terms of the information about body composition it imparts, and many papers have reported that measurements of this type fail to identify individuals with a BMI within the normal range but with elevated fat mass (J. C. K. Wells, 2019).
The noted increase in HAZ associated with having a safely managed sanitation facility provides further evidence that access to inadequate sanitation could have an impact on nutritional status (Momberg et al., 2020). This is likely to be through the direct risk of morbidity; however, while not explored here, further investigation would be needed to confirm this through pathway analyses such as structural equation modeling. Given the significance of sanitation, during a period when infants are not themselves using sanitation facilities, a point of reflection that emerges is whether this association is driven through maternal and caregiver practices and health. Untangling the effects of these different measures is the focus of the cited publication which explains why, in this manuscript, discussion around potential confounding factors like household wealth index or the hygiene index is limited (Momberg et al., 2020).
The interaction between sanitation and infection is, therefore, important; environments with poor hygiene and sanitation increase the chances of contracting an infection simply by exposing the infant to harmful pathogens. Subclinical infections have been shown to have a cumulative effect on metabolic function and growth (Dewey & Mayers, 2011). For example, environmental enteropathy has been implicated in reduced gut function, characterized by villi atrophy, crypt hyperplasia, enterocyte dysregulation, and inflammation of the lamina propria, ultimately causing an increase in intestinal permeability, allowing bacterial transfer from the gut to the blood, which contributes to chronic inflammation . In addition, there can be a direct loss of nutrients in the gut via increased intestinal  ) permeability and increased basal metabolic rate from fever can also result in loss of nutrients (Bhutta, 2006), while transport of nutrients to tissues can be impaired by infection (Stephensen, 1999). This has, in part, been explored in the South African context (Momberg et al., 2020;Said-Mohamed et al., 2019;Voth-Gaeddert et al., 2019); however, further investigation is required to better understand pathways that link, morbidity, gut function, inflammation, nutritional status, and body composition.
Another possible consideration could relate to the type of FM affected by morbidity. In infants, there is the ratio of brown to white fat to consider, as depletion of each compartment could have different implications for development due to their different roles in homeostasis (Gilsanz et al., 2013). In addition, there is also the notion of localisation of fat loss, where in cases of severe-acute undernutrition, for instance fat distribution was noted to be more central, mainly due to the loss of peripheral fat (J. C. K. Wells, 2019).
In light of the evidence raised in this study, traditional markers of adiposity such as WHZ have emerged as inadequate, and therefore, where possible the use of a more sensitive body composition indicators such as FM and FMI are critical for the investigation of the impact of infection on body composition. The authors are in the process of drafting a paper that specifically looks at various indicators for growth, body composition, and nutritional status. The limitations of z-scores are, therefore, being addressed in significant detail in forthcoming publications. From a public health perspective, these results imply that it is important to intensify efforts to prevent infection in infants in the first 6 months of life due to impacts on body composition and potentially future health. Results indicate that this prevention, relates especially to having access to a safely managed sanitation facility.
This study, like all investigations of complex interactions, has a number of limitations. The small sample size and relatively high levels of attrition are key limiting factors to the analyses performed. This was largely driven by cultural dynamics in terms of in-and out-migration of the child, where the child moved relatively frequently between caregivers and extended family in the urban Soweto setting and rural context (Ginsburg et al., 2009;Hall et al., 2018;Said-Mohamed et al., 2015).
Certain desirable covariates unfortunately had to be excluded. Exclusive breastfeeding was not included due to the fact none of the infants were exclusively breastfed at 6 months. Vaccination status w data in clinic records pertaining to vaccination status were incomplete. Another limitation is that the morbidity index consisted of potentially disparate symptoms that were summed   together. Therefore, it is not possible to distinguish which symptoms are associated with which body composition outcome. While validated against clinic records, the symptoms included do not directly relate to clinical diagnoses. While a symptom may be noted by the mother/ caregiver, the lack of or incomplete nature of the clinic records assessing the symptom in question limits our ability to ascribe them to a specific diagnosis. Notwithstanding the small sample size, the fact that associations were detected allows the authors to postulate that these associations would be supported in a larger sample. Despite sample sizes being limited in-lieu of collecting more routine and detailed data, home-visits allowed the study team to limit biases in participant answers by probing and having first-hand observation of the household.
The stringent inclusion criteria for the study helped to reduce confounding factors, thereby facilitating reliable associations. This study contributes to the limited body of knowledge assessing the effect of morbidity on infant body composition between birth and 6 months postnatally, and as far as the authors can tell, is the first, in the South African setting, specifically aimed at providing detailed information about this relationship.
The results from this study are not likely to be generalisable to other populations because of the inclusion criteria for the study. In addition, there are naturally varied responses to infection across populations due to genetic and adaptive differences, along with different diets, infection prevalence and body composition (Horns & Hood, 2012;Norgan, 1994). Therefore, studies investigating the impact of infection on body composition in other settings would be useful.

| CONCLUSION
This study brings evidence highlighting that an increase in morbidity between birth and 6 months postnatally is associated with a decrease in FM and FMI, with a significant lack of association between infection and HAZ, and WHZ. Increased HAZ in the context of a safely managed sanitation facility supports evidence that access to inadequate sanitation could have an impact on nutritional status and body composition. Reduction in FM and FMI and exposure to inflammatory cytokines associated with mounting an immune response could alter phenotypic trajectories of infants during to this period of plasticity.
In the broader context of public health, decreased FM and FMI are important due to implications for future growth, development, and health.