Age-related changes to environmental exposure: variation in the frequency that young children place hands and objects in their mouths

  • A Correction to this article was published on 27 August 2019


Children are exposed to environmental contaminants through direct ingestion of water, food, soil, and feces, and through indirect ingestion owing to mouthing hands and objects. We quantified ingestion among 30 rural Bangladeshi children < 4 years old, recording every item touched or mouthed during 6-hour video observations that occurred annually for 3 years. We calculated the frequency and duration of mouthing and the prevalence of mouth contacts with soil and feces. We compared the mouthing frequency distributions to those from US children to evaluate the appropriateness of applying the US data to the Bangladeshi context. Median hand-mouthing frequency was 43–72 times/h and object-mouthing frequency 17–34 times/h among the five age groups assessed. For half of the observations, > 75% of all hand mouthing was associated with eating. The frequency of indoor hand mouthing not related to eating was similar to the frequency of all indoor hand mouthing among children in the United States. Object-mouthing frequency was higher among Bangladeshi children compared with US children. There was low intra-child correlation of mouthing frequencies over our longitudinal visits. Our results suggest that children’s hand- and object-mouthing vary by geography and culture and that future exposure assessments can be cross-sectional if the goal is to estimate population-level distributions of mouthing frequencies. Of all observations, a child consumed soil in 23% and feces in 1%.


A primary pathway of children’s exposure to contaminants in the environment is indirect ingestion through mouthing contaminated hands and objects. A study in Tanzania that used drinking water and hand contamination to model exposure to feces in the domestic environment found that children placing contaminated hands in their mouths accounted for 97% of the total quantity of ingested fecal matter; in contrast, direct consumption of contaminated drinking water accounted for only 3% [1]. Ingestion of contaminated dust and soil through hand-to-mouth contacts has also been cited as a primary route of children’s exposure to lead and pesticides [2, 3].

Studies evaluating the risks that children in low-income countries face from direct and indirect exposure to environmental contaminants have relied on exposure data from children in the United States [1, 4]. However, if children in low-income countries have different rates of exposure than children in the United States, then employing United States exposure data to model the risk for children in low-income countries may inaccurately characterize their risk. A few studies have characterized child exposure to environmental contaminants in low-income countries [5,6,7,8,9], only two of which quantified hand-to-mouth and object-to-mouth contacts [6, 8]. Both of these studies, one in Bangladesh [8] and the other in Zimbabwe [6] used structured observation to quantify mouthing among rural children 3–18 months old. The results for the Zimbabwe data are frequency per observation, not per hour, and therefore not comparable to the US or Bangladeshi data. It is unknown why the rate of hand- and object-mouthing were higher among children in the Bangladesh than children in the United States and if these higher rates persist among older children.

Children in low-income countries may also be exposed to environmental reservoirs that are more highly contaminated or more accessible than typically found in high-income settings. For example, soils may be contaminated with lead in communities that recycle lead-acid batteries [10] or with pesticides in agricultural communities [11]. Children may also be exposed to arsenic in water [12] or mycotoxins in food [13]. In settings where families keep domestic livestock or use animal manure for cooking fuel, children may also be exposed to fecal contamination through contact with animal feces [14]. Although this paper does not address specific contaminants and environmental concentrations, a thorough characterization of children’s interactions with the environment, including these potentially highly contaminated reservoirs, will support more accurate assessment of their health risks.

Children’s interactions with the environment and their patterns of mouthing hands and objects are likely to vary with age. Cross-sectional analysis of children in the United States indicates that both indoor and outdoor hand-to-mouth contacts decrease with child age [15]. Among children, 3 months to 6 years old, object-to-mouth contact frequency is highest among children 6–23 months old [16]. This decrease in hand- and object-mouthing frequency with age may indicate that population-level exposure decreases with age. However, within a population, there may be a subset of children at higher-than-normal risk because they chronically demonstrate high-frequency mouthing behaviors [17]. A longitudinal study could determine the prevalence of high-frequency mouthing and identify the characteristics of high-frequency mouthers to consider how their exposure might be reduced [18]. A longitudinal examination of within-child correlation of mouthing behaviors over time could also inform the structure of future observational studies characterizing mouthing. If high correlation is observed, then one observation could effectively capture a given child’s average behavior across time. If the correlation is intermediate, then several observations would be required to capture the variation characterizing the child’s behavior. Alternatively, if the correlation across time is low, then even many observations would be unable to characterize a child’s “typical” behavior over the study period.

Studies quantifying the contact between children’s mouths and hands or objects have employed different methods. Hypothesizing that hand- and object-mouthing would be frequent while eating and considering that caregivers are often involved in child feeding, we sought to avoid underestimating exposure by capturing contacts both during feeding and with hands other than the child’s own. This is in contrast to several prior studies, which defined exposure time as the time a child was awake but not eating; six studies did not quantify contacts with food or other individuals’ hands [19,20,21,22,23,24], whereas three studies did not quantify contact with food but did consider contacts between the child’s mouth and hands of individuals other than the child [25,26,27]. The potential for food and food-related objects to contain contaminants of interest was taken into account in several studies. Five studies recorded contacts with food and contacts between the child’s mouth and the child’s own hand or body, but not between the child’s mouth and other individual’s hands and bodies [22, 23, 28,29,30]. Authors of two studies included feeding events and contacts between the child’s mouth and any individual’s hands [31, 32].

To complement an earlier structured observation of these children that examined the frequency of hand- and object-mouthing behaviors at one point in time [8], this study seeks to contribute to the literature by using video observation to determine the frequency and duration of environmental contacts among young children in rural Bangladesh as they aged. We use the detail captured in these video observations to evaluate potential reasons for differences in mouthing frequencies measured among children in Bangladesh and the United States, and to assess within-child correlations in mouthing behaviors over repeated visits.


Participating households were enrolled in a randomized controlled trial of water, sanitation, hygiene, and nutrition interventions (WASH Benefits trial) in rural Bangladesh [33]. In April and May 2014, we conducted an initial structured observation of hand- and object-mouthing on a subset of 150 children enrolled in the trial [8]. From these 150 children, we used simple random sampling to select 30 children to participate in three subsequent rounds of video observation from 2014 to 2016. The sample size was the mean of prior video observations [20, 21, 34,35,36,37,38, 22, 24, 25, 28,29,30,31,32]. Each of these video observations were ~ 1 year apart, with the first video observation occurring 1–2 months after the structured observations. The full set of findings from the structured observations are reported elsewhere [8]; here, we report findings from a longitudinal analysis of the three rounds of video observations.

Trained staff video recorded children’s activities at their home for 6 h during the daytime, varying the start time from 7 am to noon to capture variation in household and child activities. Observations were spread across all days of the week. Observers did not record while the child was breastfeeding or sleeping. For the video observation, video translation staff used a modified version of the Virtual Timing Device Software [39] to record hand-to-mouth, object-to-mouth, and hand-to-object contacts. Each individual contact was recorded: if the hand or mouth touched an object, stopped touching the object (either to touch another object or the air), then touched the object again, this was recorded as two separate contacts with the object. There was no minimum duration required between contacts to record them as separate contacts. Items were grouped into several categories: hands/skin (the child’s own or another individual’s), types of animals and animal feces, soil, a variety of foods, sources of water, and objects made of plastic, cloth, plant material, paper, metal, wood/brick, or other material (SI Fig. 1). An instance in which a child’s hand contacted her mouth while placing food in the mouth was counted as one hand-to-mouth and one food-to-mouth contact. Staff began translating the videos after achieving high intercoder reliability (Krippendorf’s alpha > 0.90) during training. Intercoder reliability, which considered the type of contact (mouthing/touching), child location (indoor/outdoor), hand (right/left), and item category, was assessed through the translation process, with retraining staff and re-translating observations if Krippendorf’s alpha was < 0.70.

Using this second-by-second record of children’s contacts, we calculated the frequency and duration of mouthing contacts for each child during each round of observation. We also recorded whether a hand-to-mouth contact was associated with feeding (within 15 seconds of ingesting or touching food) and recorded the number of times a hand contacted non-food items between successive hand-to-mouth contacts while eating. After classifying children into five age groups (3–5, 6–11, 12–23, 24–35, and 36–47 months of age) based on US EPA exposure assessment guidelines [40], we calculated summary frequencies and durations of contact by age group. We do not report results for children 3–5 months old since there were only two children in this age group, though we do include these children in the longitudinal analysis. To incorporate the data into probabilistic exposure models, we created age group-specific Weibull distributions of the frequency of hand-to-mouth and object-to-mouth contacts using the R package fitdistrplus. We used the Weibull distributions to compare the mouthing frequencies distributions among children in Bangladesh and the United States. The data on children in the United States are from a meta-analysis of nine studies on hand-to-mouth contacts [15] and seven studies on object-to-mouth contacts [16]. In these studies, observations lasting 1–6 h were conducted by trained observers or recorded on video in urban, suburban, and rural communities across the United States. In this US data set, mouthing frequencies were calculated after excluding time when a child was sleeping; some studies also excluded time that the child was eating [19, 41]. The US data on hand mouthing are not segregated into contacts associated with eating or not [15].

We used a linear mixed-effect model with the longitudinal data to assess the relationship between mouthing frequency and child attributes. Frequency of hand- or object-mouthing was the dependent variable, with age in months, observed mobility (immobile, crawling, or walking), sex, and location (indoor or outdoor) as independent fixed effects. We also included household ID as an independent random effect to account for household-level differences (e.g., child-rearing practices and availability of objects with which children may interact).

In addition, we evaluated the appropriateness of cross-sectional vs. longitudinal study designs for analysis of children’s exposure by examining the degree of within-child correlation between multiple rounds of observation. To determine whether some children were at potentially elevated risk from environmental contaminants owing to chronically higher exposure, we compared the distribution of mouthing frequencies for each child. We used the Kruskal–Wallis test to assess if the distribution of contact frequencies of a given child was significantly higher than that of any other child. We also assessed if any children were high-frequency mouthers, as defined by occurring in the top tertile of mouthing frequencies in all observations.


Field staff video recorded 30 children in May to June 2014 and 27 of the same children in April to May 2015 and March to May 2016, yielding a total of 84 observations. Although observations took place for 6 h, children were only recorded when awake and not breastfeeding, for an average of 3 h 41 min and 319 h of total recording (Table 1). Children spent an average of 54% of their observed time outdoors.

Table 1 Characteristics of children and households that participated in the video observations

Hand mouthing

Combining indoor and outdoor observations, median frequency of hand mouthing (own or caretaker’s) for children 6–11 months old was 43 times/h. Hand mouthing was 47 times/h for children 12–23 months old, 51 times/h for children 24–35 months old, and 72 times/h for children 36–47 months old (Fig. 1, panels 1–3; hand- and object-mouthing bootstrapped Weibull distribution parameters are given in SI Table 1; age-specific results based on the original data in SI Table 2; age- and location-specific results in SI Table 3; and mobility-specific results in SI Table 4). For every age group, the median frequency of total hand-to-mouth contacts was higher than among children in the United States [15, 16]. This difference could be accounted for by the high frequency of hand-to-mouth contacts during eating observed in Bangladesh (Fig. 2).

Fig. 1

Frequency of hand-to-mouth contacts not related to eating, hand-to-mouth contacts related to eating, all hand-to-mouth contacts, and object-to-mouth contacts among young children in rural Bangladesh

Fig. 2

Probability distributions for the frequency (contacts/h) of indoor hand-to-mouth contacts (top) and object-to-mouth contacts (bottom) among children in rural Bangladesh and United States. In the plot for children 36–47 months old, the US data include children up to 60 months old [15]. The 95% confidence interval is not available for the US data

Neither age, child sex, or indoor/outdoor location of play were significantly related to the frequency of children’s total, dietary, or non-dietary hand-to-mouth contacts. Mobility was not associated with the frequency of all hand mouthing or hand mouthing while eating but was associated with a decrease in hand mouthing unrelated to eating. Transitioning from crawling to walking corresponded to 24 fewer hand-to-mouth contacts per hour awake and not eating (p < 0.01).

For all ages combined, the mean duration of each hand contact with the mouth when eating was 3.5 s (standard deviation (sd) = 5.1 s) (SI Table 5). Neither age nor mobility were associated with the duration of mouthing hands while not eating, but mobility and location were associated with the duration of hand mouthing while eating. Transitioning from crawling to walking was associated with a decrease in duration of each contact by 0.7 s, while transitioning from inside to outside was associated with an increase in duration of each contact by 0.6 s.

Mouthing associated with eating

Children consumed a variety of foods by hand and with utensils. The average percentage of the observation time that children consumed food other than breastmilk varied from 13% for children 6–11 months old to 19% for children 36–47 months old (Fig. 3, panel 1). Hand-to-mouth contacts while eating accounted for > 75% of all hand-to-mouth contacts in half of all children observed (Fig. 3, panel 2). The median duration of all children’s eating periods was 42 s. The frequency that caregivers fed children by hand significantly decreased with child age (p < 0.01). On average, caregivers were involved in 47% of food-related hand-to-mouth contacts among children 6–11 months old, compared with 23% for children 12–23 months old and 12% for children 24–36 months old (Fig. 3, panel 3). Foods that were commonly touched and consumed were fresh fruit (29% of all hand-to-food contacts), rice or a mixture of rice and lentils (21%), dry foods such as puffed rice (16%), and foods from packages, such as chips (10%) (SI Fig. 2). Although children in Bangladesh commonly eat solid foods with their hands, a median of 20% of all object-to-mouth contacts were with utensils (Fig. 3, panel 4). Of the 730 eating periods in which children had hand-to-mouth contact while eating, the child contacted the soil once during 20 eating periods (2.7%), more than once during 18 eating periods (2.5%), and never in the other periods.

Fig. 3

Percentage of hand-to-mouth contacts that occurred during eating or involved other individuals and the percentage of object-to-mouth contacts that involved utensils among young children in rural Bangladesh. The dark middle lines denote the median, the edges of the boxes mark the 75th and 25th percentile, and the whiskers extend to values ± 1.5 times the interquartile range

Touching and mouthing objects

Median object-mouthing frequency among children 34 times/h for children 6–11 months old, 22 times/h for children 12–23 months old, 19 times/h for children 24–35 months old, and 17 times/h for children 36–47 months old (Fig. 1, panel 4; bootstrapped Weibull distribution parameters are given in SI Table 1; age-specific results based on the original data in SI Table 2; age- and location-specific results in SI Table 3; and mobility-specific results in SI Table 4). For all age groups, the median frequency of object-to-mouth contacts was higher among children in this study compared with children in the United States (Fig. 2).

Children touched numerous types of objects including their mother’s clothes (33% of total contacts), plastic objects, often used as toys (31%), and plant material such as leaves and sticks (12%). Items that children put into their mouths included plastic objects (51% of all object-to-mouth contacts), metal (12%), plant material (9%), and cloth (9%). As with hand mouthing, neither age, child sex, nor indoor/outdoor location of play were significantly related to the frequency that children mouthed or touched objects.

Mobility was significantly associated with the frequency of mouthing and touching objects. Transitioning from crawling to walking was associated with a decrease in mouthing objects by 30 contacts/hr and a decrease in touching objects by 51 contacts/hr. Being outside was associated with a decrease in touching objects by 72 contacts/hr.

For all ages combined, the mean duration of each object-to-mouth contact was 7.7 s (sd = 13.7 s) (SI Table 5). Each hand-to-object contact averaged 10.0 s (sd = 16.4 seconds). Aging 1 year was associated a 0.8-second increase in the duration of mouthing an object and 1.2-second increase in the duration of touching an object. Walking was associated with touching contacts that were 2.4 s longer than touching contacts among crawlers. Being outdoor was associated with touching contacts that were 0.73 s/contact longer.

Contact with soil, animal feces, and water

The prevalence and frequency of observed soil and feces contact with children’s mouth was low (Table 2). There was a total of 20 soil ingestion events during the video observations. In eight instances, the ingested soil was from the household floor, and in 12 instances the soil ingestion occurred outside. Of the 15 children who mouthed soil at least once, one child mouthed soil in each of the three observations, two children in two of three observations, and one child twice in the same observation; the others mouthed soil only once. Approximately one-third of children 6–11 and 12–23 months old were observed placing soil in their mouths or licking hard-packed earth (Table 2). In contrast, 7% of children 24–35 months old, and no children 36–47 months old were observed directly ingesting soil. Six of the 20 soil ingestion events were by children who could only crawl, whereas 14 ingestion events were by children who could walk. The soil mouthing frequency was not associated with mobility (p = 1.00).

Table 2 Prevalence and frequency of observed soil (top) and feces (bottom) ingestion during observations of young children

Hand contact with soil was highest among children 6–11 months old, who had a median contact frequency of 124 contacts/h. Aging 1 month was associated with a 3.5 fewer hand-touching contacts with soil every hour, starting from a theoretical 125 contact/h for a child of 0 months (p < 0.001) (Fig. 4, panel 4). Likewise, increase in mobility from crawling to walking was associated with a significant decrease in the frequency of touching soil, by 53 contacts/h (94 vs. 41 contacts/h; p = 0.01).

Fig. 4

Frequency of touching water, foods, soil, and objects among young children in rural Bangladesh

Animal feces were observed in 70–100% of households during each round of observation (Table 1) and a total of eight children touched their own feces, other human feces, or the droppings of chickens, goats, or cows. During all the video observations, there was a single instance in which a child placed feces in his mouth (Table 2). This child walked over to a pile of dry cow dung chips stored for cooking fuel and put a small piece into his mouth. Without any prompting from the videographer, the caregiver immediately rushed over to the child and used her finger to sweep the feces out of his mouth. There were nine observations of children touching animal feces; these contacts were more common among children who were observed when they could crawl (6 of 38 children; 15%) compared both with children who could not yet crawl (0/8; 0%) and children who were already walking (4/71; 6%). The video observations captured 132 contacts with animals, including contacts with goats and cows (51%) and chickens (48%). In addition, children of all ages had contact with water stored in the household, tubewell water, and water that was being used for a specific purpose, such as washing hair or feeding animals, but did not contact pond water.

Intra-child variation in mouthing frequencies

The longitudinal study design revealed low intra-child correlation over time. Among children who were observed walking, intercluster correlation was 0.11 for mouthing hands while not eating, − 0.06 for mouthing hands while eating, and 0.19 for mouthing objects. Children were observed at most once while they were immobile or crawling, so intra-child variability could not be assessed in these mobility groups. Within-child correlation was lower when mobility was not considered.

The distribution of children’s mouthing frequencies (up to three timepoints per child) did not significantly differ by child (p = 0.32 for hands and p = 0.89 for objects). Furthermore, there was no child whose mouthing frequency was in the top tertile of frequencies in all three observations.


As children aged, the frequency of hand mouthing while not eating tended to decrease, whereas the frequency of hand mouthing while eating tended to increase, but in this sample neither was significantly associated with child age. Small sample size within each age group, wide variation in mouthing frequencies at any given age, and a short interval of comparison (0–47 months) may all contribute to the lack of associations with age. However, walking was associated with a decrease in hand mouthing while not eating compared with crawling, indicating that milestones in gross motor development, rather than age in months, may be more suitable for categorizing children by mouthing exposures.

A prior analysis of these children when they were 3–18 months old found that their frequencies of hand- and object-mouthing were higher than their same-age USA peers [8]. This study finds that this difference extends to children < 48 months old and can be accounted for by Bangladeshi children’s hand-to-mouth contacts that occur while eating. Different hand-mouthing frequencies between groups of children may or may not create substantial differences in their oral exposure (direct and indirect ingestion). The extent to which hand-to-mouth contacts affects risk of ingesting contaminants depends on the degree of recontamination between each hand-to-mouth contact. In this study, we observed children’s hands contacting soil in 5% of eating events. Assuming that there is no recontamination between hand-to-mouth contacts while eating and the majority of contamination is transferred during the first hand-to-mouth contact of each eating event [42] would be one extreme scenario. In this case, risk assessments should consider the frequency of eating events, rather the frequency of individual hand-to-mouth contacts while eating. Assuming that there is recontamination between every contact, as we assume for hand-to-mouth contacts not related to eating, would result in the opposite extreme. In this case, risk assessments should consider the frequency of hand mouthing that occurs for any reason. Sensitivity analysis of the frequency and degree of recontamination could provide a quantitative estimate of how exposure differs under these scenarios.

Object mouthing was substantially more frequent among children in Bangladesh than children in the United States [15, 16]. Although the level of contamination on various objects likely varies substantially, if contamination of objects is an important pathway for contaminant transmission, then using data from US children rather than Bengali children would underestimate Bangladeshi children’s environmental risks. Further research on the contaminant levels on objects that children mouth could help clarify the risks from object-mouthing exposures.

In addition to mouthing and touching objects more frequently, the children in our study contacted objects that may be highly contaminated (such as soil and feces) and that are likely less prevalent in the context of children in high-income countries or urban areas. For example, bare soil was present in both children’s indoor and outdoor environments in our study setting and placing soil in the mouth was common. For children 6–11 months old, the prevalence of placing soil into the mouth (42%), 8–15 percentage points higher than in prior observations of children in rural Bangladesh and Ghana. For children 12–23 months old, the soil ingestion prevalence of 26% that we observed was within the 18–38% range reported in studies in Bangladesh, Ghana, and Zimbabwe [6,7,8, 43] (Table 2). This supports the growing evidence that direct ingestion of soil may contribute to enteric dysfunction and illness [7, 43].

Children who consistently mouth hands or objects more than other children may share characteristics that define them as a distinct subpopulation that may be at elevated risk from environmental contamination. If children in this subpopulation can be readily identified, they could be targeted for surveillance or intervention. For example, autistic children have demonstrated a higher prevalence of pica compared with non-autistic children [44]. Their higher rates of pica may be associated with elevated blood lead levels [17], indicating that these children may benefit from blood lead screening and, if high levels are found, interventions to reduce pica. We observed low intra-child correlation of mouthing frequencies in this longitudinal study. Of the 30 study children, none mouthed hands or objects with consistently higher frequency than other children. If there is a subpopulation of children who are high-frequency mouthers and at high risk from environmental contamination owing to elevated exposure, our findings suggest that they are a relatively small proportion of the population (< 6%, see SI for details).

It is unclear how contact duration affects the efficiency of contaminant transfer. One study demonstrated that hand-to-mouth transfer of riboflavin, a surrogate for pesticides, was independent of whether each contact was 2 s or 20 s, but this may have been due to the overwhelming effects of other test parameters [45]. Using the range of durations identified in this study, further studies could be done to elucidate the effect of duration on transfer efficiency of various contaminants. If duration is found to substantially moderate transfer efficiency, future studies could also collect detailed contact duration data using video or structured observations processed with VideoTraq or LiveTrak software [46]. These durations could then be used to more accurately estimate exposure.

Climactic factors may also influence mouthing. Our observations were conducted during the hot and dry period of March to June, when children spent half of their time outdoors. In seasons that are cooler or have more rain, children may spend more or less time outside and/or interact with different types of objects. However, if the young children in our study placed their hands and objects in their mouths as a form of exploration [47], then their frequency of mouthing may be independent of season. This hypothesis is supported by the finding that their mouthing frequencies were not significantly different indoors compared with outdoors.

Our assessment of children’s mouthing and touching behaviors at four timepoints over a 3-year period has several implications for accurately estimating young children’s exposure to environmental contaminants through indirect ingestion pathways. For half of the children in this study, hand mouthing that occurred while eating accounted for the three-quarters of all hand-to-mouth contacts. This suggests that in similar settings, where children that eat by hand because of age, culture, or the types of food they commonly consume, hand mouthing while eating may contribute > 75% of all contamination transferred via hand mouthing. Studies that neglect exposure that occurs while eating may thus substantially underestimate risk. However, if the degree of hand recontamination while eating is low, as observed among children in this study, considering the frequency of eating events rather than frequency of individual hand-to-mouth contacts while eating may result in more-accurate exposure assessment. Second, low intra-child correlation in mouthing frequencies and the association of frequencies with child mobility, rather than age, implies that future exposure assessments can be cross-sectional. In studies of children who do not yet attend school, which would likely modify their behaviors, children could be stratified by mobility rather than narrow age bands.

Change history

  • 27 August 2019

    An amendment to this paper has been published and can be accessed via a link at the top of the paper.


  1. 1.

    Mattioli MCM, Davis J, Boehm AB. Hand-to-mouth contacts result in greater ingestion of feces than dietary water consumption in Tanzania: a quantitative fecal exposure assessment model. Environ Sci Technol 2015;49:1912–20.

  2. 2.

    Laidlaw MAS, Filippelli GM. Resuspension of urban soils as a persistent source of lead poisoning in children: a review and new directions. Appl Geochem 2008;23:2021–39.

  3. 3.

    Guney M, Zagury GJ, Dogan N, Onay TT. Exposure assessment and risk characterization from trace elements following soil ingestion by children exposed to playgrounds, parks and picnic areas. J Hazard Mater. 2010;182:656–64.

  4. 4.

    Wang Y, Moe CL, Null C, Raj SJ, Baker KK, Katharine A, et al. Multipathway quantitative assessment of exposure to fecal contamination for young children in low-income urban environments in Accra, Ghana: the SaniPath analytical approach. Am J Trop Med Hyg 2017;97:1009–19.

  5. 5.

    Marquis GS, Ventura G, Gilman RH, Porras E, Miranda E, Carbajal L, et al. Fecal contamination of shanty town toddlers in households with non-corralled poultry, Lima, Peru. Am J Public Health 1990;80:146–9.

  6. 6.

    Ngure FM, Humphrey JH, Mbuya MNN, Majo F, Mutasa K, Govha M, et al. Formative research on hygiene behaviors and geophagy among infants and young children and implications of exposure to fecal bacteria. Am J Trop Med Hyg 2013;89:709–16.

  7. 7.

    George C, Oldja L, Biswas S, Perin J, Lee G, Kosek M, et al. Geophagy is associated with environmental enteropathy and stunting in children in rural Bangladesh. Am J Trop Med Hyg. 2015;92:1117–24.

  8. 8.

    Kwong LH, Ercumen A, Pickering AJ, Unicomb L, Davis J, Luby SP. Hand- and object-mouthing of rural bangladeshi children 3–18 months old. Int J Environ Res Public Health 2016;13:563–81.

  9. 9.

    Teunis PFM, Reese HE, Null C, Yakubu H, Moe CL. Quantifying contact with the environment: behaviors of young children in Accra, Ghana. Am J Trop Med Hyg 2016;94:920–31.

  10. 10.

    Haefliger P, Mathieu-Nolf M, Lociciro S, Ndiaye C, Coly M, Diouf A, et al. Mass lead intoxication from informal used lead-acid battery recycling in Dakar, Senegal. Environ Health Perspect 2009;117:1535–40.

  11. 11.

    Cantu-Soto EU, Meza-Montenegro MM, Valenzuela-Quintanar AI, Félix-Fuentes A, Grajeda-Cota P, Balderas-Cortes JJ, et al. Residues of organochlorine pesticides in soils from the southern Sonora, Mexico. Bull Environ Contam Toxicol 2011;87:556–60.

  12. 12.

    Mukherjee A, Sengupta MK, Hossain MA, Ahamed S, Das B, Nayak B, et al. Arsenic contamination in groundwater: a global perspective with emphasis on the Asian scenario. J Heal Popul Nutr 2006;24:142–63.

  13. 13.

    Bankole S, Schollenberger M, Drochner W. Mycotoxins in food systems in Sub Saharan Africa: a review. Mycotoxin Res 2006;22:163–9.

  14. 14.

    Ercumen A, Pickering AJ, Kwong LH, Arnold BF, Parvez SM, Alam M, et al. Animal feces contribute to domestic fecal contamination: evidence from E. coli measured in water, hands, food, flies, and soil in Bangladesh. Environ Sci Technol 2017;51:8725–34.

  15. 15.

    Xue J, Zartarian V, Moya J, Freeman N, Beamer P, Black K, et al. A meta-analysis of children’s hand-to-mouth frequency data for estimating nondietary ingestion exposure. Risk Anal. 2007;27:411–20.

  16. 16.

    Xue J, Zartarian V, Tulve N, Moya J, Freeman N, Auyeung W, et al. A meta-analysis of children’s object-to-mouth frequency data for estimating non-dietary ingestion exposure. J Expo Sci Environ Epidemiol. 2010;20:536–45.

  17. 17.

    Cohen DJ, Johnson WT, Caparulo BK. Pica and eleveated blood lead levels in autistic and atypical children. Am J Dis Child 1976;130:47–8.

  18. 18.

    Thompson KM. Changes in children’s exposure as a function of age and the relevance of age definitions for exposure and health risk assessment. MedGenMed. 2004;6:2.

  19. 19.

    Tulve NS, Suggs JC, McCurdy T, Cohen Hubal Ea, Moya J, Hubal AC, et al. Frequency of mouthing behavior in young children. J Expo Anal Environ Epidemiol. 2002;12:259–64.

  20. 20.

    AuYeung W, Canales RA, Beamer P, Ferguson AC, Leckie JO. Young children’s mouthing behavior: an observational study via videotaping in a primarily outdoor residential setting. J Child Heal 2004;2:271–95.

  21. 21.

    Black K, Shalat SL, Freeman NC, Jimenez M, Donnelly KC, Calvin JA. Children’s mouthing and food-handling behavior in an agricultural community on the US/Mexico border. J Expo Sci Environ Epidemiol. 2005;15:244–51.

  22. 22.

    Brinkman S, Gialamas A, Jones L, Edwards P, Maynard E. Child activity patterns for environmental exposure assessment in the home. Adelaide, Australia; 1999.

  23. 23.

    Greene MA Oral DINP Intake Among Young Children. Bethesda: U.S. Consumer Product Safety Commission; 2002.

  24. 24.

    Reed KJ, Jimenez M, Freeman CG, Lioy PJ, Freeman NC, Lioy PJ. Quantification of children’s hand and mouthing activities through a videotaping methodology. J Expo Anal Environ Epidemiol 1999;9:513–20.

  25. 25.

    Freeman N, Jimenez M, Reed K, Gurunathan S, Edwards R, Roy A, et al. Quantitative analysis of children’s microactivity patterns: the Minnesota Children’s Pesticide Exposure Study. J Expo Anal Environ Epidemiol. 2001;11:501–9.

  26. 26.

    Groot M, Lekkerkerk M, Steenbe. Mouthing behaviour of young children: asn observational study. Wageningen, The Netherlands: Agricultural University Wageningen, 1998.

  27. 27.

    Smith SA, Norris B. Reducing the risk of choking hazards: mouthing behavior of children aged 1 month to 5 years. Inj Control Saf Promot. 2003;10:145–54.

  28. 28.

    Zartarian VG, Ferguson AC, Leckie JO. Quantified dermal activity data from a four-child pilot field study. J Expo Anal Environ Epidemiol 1997;7:543–53.

  29. 29.

    Ko S, Schaefer PD, Vicario CM, Binns HJ, Safer FORTHE, Project Y. Relationships of video assessments of touching and mouthing behaviors during outdoor play in urban residential yards to parental perceptions of child behaviors and blood lead levels. J Expo Sci Environ Epidemiol. 2007;17:1747–57.

  30. 30.

    Tsou M, Özkaynak H, Beamer P, Dang W, Hsi H-C, Jiang C-B, et al. Mouthing activity data for children aged 7 to 35 months in Taiwan. J Expo Sci Environ Epidemiol 2015;25:388–98.

  31. 31.

    Beamer P, Key ME, Ferguson AC, Canales RA, Auyeung W, Leckie JO. Quantified activity pattern data from 6 to 27-month-old farmworker children for use in exposure assessment. Environ Res 2008;108:239–46.

  32. 32.

    Beamer PI, Luik CE, Canales Ra, Leckie JO. Quantified outdoor micro-activity data for children aged 7–12-years old. J Expo Sci Environ Epidemiol 2012;22:82–92.

  33. 33.

    Arnold BF, Null C, Luby SP, Unicomb L, Stewart CP, Dewey KG, et al. Cluster-randomised controlled trials of individual and combined water, sanitation, hygiene and nutritional interventions in rural Bangladesh and Kenya: the WASH Benefits study design and rationale. BMJ Open 2013;3:e003476.

  34. 34.

    Ruff HA, Mccarton C, Kurtzberg D, Vaughan HG, Ruff HA, Mccarton C, et al. Preterm infants’ manipulative of objects exploration. Child Dev 1984;55:1166–73.

  35. 35.

    Ruff HA. Components of attention during infants’ manipulative exploration. Child Dev 1986;57:105–14.

  36. 36.

    Leckie J, Naylor K, Canales R, Ferguson A, Cabrera N, Hurtado A, et al. Quantifying children’s microlevel activity data from existing videotapes. Research Triangle Park, NC; 2000.

  37. 37.

    Freeman NC, Hore P, Black K, Jimenez M, Sheldon L, Tulve N, et al. Contributions of children’s activities to pesticide hand loadings following residential pesticide application. J Expo Anal Environ Epidemiol 2005;15:81–8.

  38. 38.

    Davis E, Cumming O, Aseyo RE, Muganda DN, Baker KK, Mumma J, et al. Oral contact events and caregiver hand hygiene: Implications for fecal-oral exposure to enteric pathogens among infants 3–9 months living in informal, Peri-Urban communities in Kisumu, Kenya. Int J Environ Res Public Health 2018;15:192.

  39. 39.

    Ferguson AC, Canales Ra, Beamer P, Auyeung W, Key M, Munninghoff A, et al. Video methods in the quantification of children’s exposures. J Expo Sci Environ Epidemiol 2006;16:287–98.

  40. 40.

    U.S. Environmental Protection Agency. Guidance on Selecting Age Groups for Monitoring and Assessing Childhood Exposures to Environmental Contaminants [Internet]. Washington, DC; 2005. Available from: EPA/630/P-03/003F.

  41. 41.

    Greene MA Mouthing Times Among Young Children From Observational Data [Internet]. Bethesda, MD; 2002. Available from:

  42. 42.

    Rodes CE, Newsome JR, Vanderpool RW, Antley JT, Lewis RG, Antley T, et al. Experimental methodologies and preliminary transfer factor data for estimation of dermal exposures to particles. J Expo Anal Environ Epidemiol 2001;11:123–39.

  43. 43.

    Bauza V, Byrne DM, Trimmer JT, Lardizabal A, Atiim P, Asigbee MAK, et al. Child soil ingestion in rural Ghana: frequency, caregiver perceptions, relationship with household floor material, and associations with child diarrhoea. Trop Med Int Health 2018;23:558–69.

  44. 44.

    Kinnell HG. Pica as a feature of autism. Br J Psychiatry 1985;147:80–2.

  45. 45.

    Cohen Hubal EA, Suggs JC, Nishioka G, Ivancic WA, Cohen Hubal EA, Suggs JC, et al. Characterizing residue transfer efficiencies using a fluorescent imaging technique. J Expo Anal Environ Epidemiol 2005;15:261–70.

  46. 46.

    Julian TR, Bustos C, Kwong LH, Badilla AD, Lee J, Bische HN, et al. Quantifying human-environment interactions using videography in the context of infectious disease transmission. Geospat Health 2018;13:195–7.

  47. 47.

    Gibson EJ. Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. Annu Rev Psychol 1988;39:1–41.

Download references


We gratefully acknowledge the participating study families, as well as collaborators responsible for video translation software, data collection and video translation: Robert Canales, Carla Bustos, Rita Chowdhury, Nisa Sultana, Eashmat Annay, Shamima Islam Mina, Nusrat Jahan, Akhe Nur Begum, and Md. Sajjad Rahman. This study was funded by the Stanford Center for Innovation in Global Health and the World Bank; the WASH Benefits Study, in which it was embedded, was funded by the Bill & Melinda Gates Foundation. This material is based upon work supported by the Stanford Wood’s Institute for the Environment Goldman Graduate Fellowship and the National Science Foundation Graduate Research Fellowship under Grant No. DGE-114747. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors(s) and do not necessarily reflect the views of the National Science Foundation.

Author information

Correspondence to Laura H. Kwong.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

Caregivers of the participating children provided informed written consent before the observations. The study was conducted in accordance with the Declaration of Helsinki and received ethics clearance from Stanford University (Protocol 25863), University of California, Berkeley (2011-09-3652), and the International Centre for Diarrhoeal Disease Research, Bangladesh (PR-11063).

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark


  • micro-level activity time series data set (MLATS)
  • exposure factors
  • indirect ingestion
  • mouthing
  • microactivities
  • Bangladesh