Maternal distraction during breast- and bottle-feeding among WIC and non-WIC mothers

Abstract

Objective

To explore the prevalence and correlates of maternal distraction during infant feeding within a sample of WIC and non-WIC mothers

Design

Mothers kept diaries of their infants' feeding patterns

Participants

Mothers (n=75) with infants ≤6 months old

Main Outcome Measures

Within the diaries, mothers recorded what else, if anything, they did during the feeding. Mothers also completed demographic, feeding styles, and infant temperament and eating behaviors questionnaires.

Analysis

Mothers’ responses were coded into thematic categories. Feedings were classified as distracted when the mothers reported doing something other than feeding and/or interacting with the infant. Logistic regression was used to explore whether mothers exhibited different levels of distraction when breast-versus bottle-feeding. Multiple stepwise regression was used to explore associations between distracted feeding and characteristics of mothers and infants.

Results

Distractions were reported during 43% of feedings; 26% involved technological distractors. Mothers who were multiparous and perceived their infants had greater appetites reported greater levels of any distractions during feeding. Mothers who were racial/ethnic minorities, adhered to laissez faire feeding style, had younger infants, and perceived their infants to have lower food responsiveness and greater appetite reported greater levels of technological distractions. WIC status was not associated with mothers’ levels of distracted feeding.

Conclusions and Implications

Mothers reported a wide variety of distractions during both breast- and bottle-feeding; higher levels of distraction were associated with characteristics of both mothers and infants. Further research is needed to understand whether and how maternal distraction impacts feeding outcomes.

INTRODUCTION

A growing body of research highlights the importance of caregiver sensitivity and responsiveness during feeding for promoting infants’ abilities to self-regulate intake¹ and healthy infant weight gain trajectories.² Additionally, caregivers’ sensitivity and responsiveness in both feeding and non-feeding contexts helps infants develop effective emotional, cognitive, and behavioral self-regulatory abilities,³ which is predictive of significantly better stress reactivity,⁴ lower risk for internalizing and externalizing problems,⁵ and lower risk of obesity and related comorbidities^6,7 during later life. Thus, it is well established that an important foundation for promoting healthy development during early childhood is the promotion of caregivers’ sensitivity and responsiveness during caregiver-infant interactions. However, the Institute of Medicine (IOM)⁸ and others⁹ have highlighted the surprising paucity of research aimed at understanding how to promote caregivers’ sensitivity and responsiveness within feeding contexts and the need for more research in this field. This is especially true for low-income and racial/ethnic minority caregivers, who are more likely to use non-responsive feeding practices^10–12 and whose infants are at higher risk for rapid weight gain and obesity.^13–15

Although there are many reasons why caregivers may exhibit low responsiveness to infant cues during feeding interactions, few studies have examined how maternal distraction (e.g., watching television [TV], using a mobile device) may impact infant feeding interactions and outcomes.^16–19 Research in older children and adults illustrates that distracted eating is a common occurrence²⁰ and that eating while engaging with technological distractions impacts meal outcomes, leading to lower awareness of satiety cues and overeating when compared to meals when no distractions are present.^21,22 Very few studies have examined whether “distracted feeding” has similar and/or unique impacts on caregiver-child feeding interactions.

Given the ever-increasing accessibility of handheld technologies and on-demand entertainment²³ - even among lower-income, minority populations²⁴ who are more likely to use mobile devices as their sole source of internet access²⁵ - combined with the large number of feedings required by young infants on a daily basis,²⁶ it is possible that many caregivers regularly attend to distractors while feeding their infants as a means to cope with the large volume of time they must dedicate to feeding. Indeed, in a recent study of bottle-feeding dyads, mothers reported distractions during 52% of bottle-feedings, with almost 1/3 of these feedings involving technological distractors.¹⁹ Additionally, in a laboratory-based study, mothers who were distracted while bottle-feeding their infants were significantly less sensitive to their infants’ cues than mothers who were not distracted.¹⁸ These previous studies have also illustrated that tendencies toward distracted feeding are associated with certain maternal characteristics, such as age and parity,¹⁹ and that associations between distraction and feeding outcomes may be moderated by dimensions of infant temperament,¹⁸ such as infants’ abilities to self-regulate (i.e., orienting/regulation capacities) and levels of impulsivity (i.e., surgency/extraversion).

Given these previous findings focused solely on bottle-feeding mothers, further research is needed to understand whether similar levels of distraction are present in breastfeeding dyads and whether certain mother-infant dyads are more likely than others to engage in distracted feeding. To this end, the objective of the present study was to further explore the prevalence of maternal distraction within a sample of both breast- and bottle-feeding WIC and non-WIC participants. A secondary aim of this study was to examine whether maternal (i.e., feeding mode, WIC status, education, race/ethnicity, parity, BMI, feeding styles) and infant (i.e., sex, age, weight status, temperament, eating behaviors) characteristics are associated with mothers’ tendencies toward distracted feeding. Based on findings of previous research focused on distracted feeding,^18,19 it was hypothesized that higher levels of distraction would be associated with greater maternal age and parity and lower levels of responsive feeding style. Additionally, previous research has illustrated breastfeeding mothers exhibit greater tendencies toward responsive feeding;27,28 accordingly, it was hypothesized that bottle-feeding mothers would show greater tendencies toward distracted feeding than breastfeeding mothers. Finally, given previous research illustrating the importance of maternal responsiveness for infants’ developing abilities to self-regulate, it was hypothesized that greater levels of maternal distraction would be associated with lower general (i.e., the orienting/regulation capacity dimension of temperament) and feeding specific (i.e., the food responsiveness and satiety responsiveness dimensions of infant eating behaviors) self-regulatory capacities for infants.

METHODS

Participants and Recruitment

Mothers (n=75) who participated in infant feeding studies^12,19,29 were asked to keep a diary of their infants' feeding patterns for 1–6 days (total number of records = 476; total number of recorded feedings = 2982). Feeding record data for a subset of 41 formula-feeding infants have been published previously.¹⁹ Inclusion criteria for infants were: 1) healthy, 2) born full-term, 3) between birth and 6 months of age, and 4) not yet introduced to solid foods. Inclusion criteria for mothers were: 1) between 18 and 40 years of age and 2) absence of gestational diabetes or any complications during pregnancy or birth that lead to infant feeding problems. Both breast- and formula-feeding dyads were included in the present study. Participants were recruited through fliers posted in Women, Infant & Children (WIC) offices, breastfeeding support groups, libraries, coffee shops, and local pediatric offices, as well as through targeted Facebook advertisements. Mothers were provided a $25 cash compensation for their participation. All study procedures were reviewed under expedited review and approved by the university Institutional Review Board. Both oral and written informed consent was obtained from each mother.

Procedures and Measures

Mothers received feeding records via postal mail or email. Through both verbal instruction by a research assistant and written instructions on the form, mothers were asked to record, for each feeding: 1) start and end time; 2) what was fed (e.g., formula, breast milk from the breast, breast milk from a bottle); 3) amount fed (if possible); and 4) what else, if anything, they were doing while feeding their infants.

Printed records were collected from all mothers during a laboratory visit several days later, at which time mothers also completed a demographic questionnaire. Mothers also completed the Infant Behavior Questionnaire-Revised Very Short Form (IBQ-R), which assesses infants’ levels of surgency/extraversion, orienting/regulation capacity, and negative affect (subscale scores range from 1 – 7);³⁰ the Infant Feeding Styles Questionnaire (IFSQ),³¹ which assesses mothers’ self-reported laissez-faire (example item: “I think it is okay to prop an infant’s bottle”), restrictive (example item: “It’s important for the parent to decide how much an infant should eat”), pressuring (example item: “I try to get my child to eat even if s/he seems not hungry”), and responsive (example item: “My child knows when s/he is hungry and needs to eat”) feeding styles (all subscale scores range from 1–5); and the Baby Eating Behavior Questionnaire (BEBQ),³² which assesses mothers’ perceptions of infants’ levels of enjoyment of food, food responsiveness, satiety responsiveness, slowness in eating and general appetite (subscale scores range from 1 – 5). All questionnaire subscales showed good internal consistency in this sample, with Cronbach’s alpha scores ranging from acceptable (α=.70–.79) to good (α=.80–.89): the IBQ-R subscales ranged from α=.78–83, the IFQS subscales ranged from α=.70–.85, and the BEBQ subscales ranged from α=.71–.87.

A trained research assistant collected weight and length or height measurements in triplicate for infants and mothers using an infant scale and infantometer (models 374 and 233; Seca, Hamburg, Germany), and an adult scale and stadiometer (model 736; Seca, Hamburg, Germany), respectively. Infant anthropometric data was later normalized to z-scores using the World Health Organization (WHO) Anthro software version 3.0.1 (http://who.int/childgrowth/en/). Mothers’ weight and height were used to calculate Body Mass Index (BMI; kg/m²).

Data Analysis

Mothers’ responses to the question: “what else, if anything, were you doing while feeding your infants?” were sorted into thematic categories using constant comparison within the framework of grounded theory.³³ Prior to coding, the first author developed a coding manual with an initial list of themes that were found in previous research examining the prevalence of distraction among bottle-feeding dyads.¹⁹ Three trained coders independently coded all records based on the coding manual, but also identified additional themes that were not present in the previous research. Results were then reviewed and compared for validity, and any discrepancies in theme identification or coding were discussed. Inter-coder reliability was established by comparing the common coding of a total of 10 records by all coders. The mean Pearson’s rho for the correspondence among coders was ρ(7df)>0.80, indicating good reliability between coders.

After coding was complete, themes were used to classify feedings into 2 categories: 1) mother was distracted (e.g., watching TV, using a computer) versus 2) mother was not distracted (e.g., nothing was specified, interacting with the infant). Distractions were also further classified into technological (e.g., watching TV, using a computer, smart phone, or tablet) versus not (e.g., reading, doing housework). To obtain a measure of each mothers’ intensity of distracted feeding, the percentage of feedings during which the mother reported any distraction was calculated for each mother (= [number of feedings wherein a distraction was reported/total number of feedings reported]*100). Similarly, the percentage of feedings during which the mother reported technological distractions was calculated for each mother (= [number of feedings wherein a technological distraction was reported/total number of feedings reported]*100).

All quantitative analyses were conducted using SAS v.9.4 (July 2013; SAS Institute Inc., North Carolina, USA). Descriptive statistics were calculated to summarize sample demographics, mothers’ frequency of different activities during reported feedings, and distracted versus not distracted feeding. Characteristics of WIC versus non-WIC participants were compared using Analysis of Variance (ANOVA) and Chi-squared tests. Mixed effects models were used to examine associations between mothers’ reported feeding durations and reports of any type of distraction or technological distraction during feeding, while also controlling for within-person correlation among repeated measurements of feeding durations and distractions.

To explore whether mothers exhibited different levels of distraction when breast-versus bottle-feeding, logistic regression with estimation via generalized estimating equations was used to determine whether feeding mode during each recorded feeding predicted the probability of mothers’ reports of distracted versus not distracted feeding. Logistic regression models accounted for the within-person correlation among repeated measurements for valid estimation of associations between feeding mode and types of distraction, and associated standard errors.

To explore associations between distracted feeding and characteristics of mothers (i.e., feeding mode, WIC status, education, race/ethnicity, parity, BMI, feeding styles) and infants (i.e., sex, age, weight status, temperament, eating behaviors) multiple stepwise regression was used to determine the combination of infant and maternal characteristics that best predicted mothers’ intensities of any distractions and technological distractions during feeding. Categorical variables were dummy coded prior to inclusion in regression models (i.e., feeding mode, mothers’ education level [high school degree or less versus some college or college degree], mothers’ race/ethnicity [White versus Black, Hispanic, or other]). Additionally, all assumptions for multicollinearity of predictors, homoscedasticity, normality of residuals, and linearity were assessed prior to all linear regression analyses. When applicable, results are presented as percentages (n) or means ± standard deviations (SD). A significance level of P < .05 was used to indicate significant differences.

RESULTS

Sample Characteristics

Table 1 summarizes sample characteristics for WIC versus non-WIC mothers. As illustrated in Table 1, WIC infants were significantly younger and had significantly higher weight-for-length z-scores than non-WIC infants. WIC mothers were significantly younger, had significantly higher BMI, and had more children than non-WIC mothers. Similar to national data on characteristics of WIC participants,³⁴ WIC mothers reported significantly lower family incomes and education levels than non-WIC mothers and greater proportions of WIC mothers were racial/ethnic minorities and were formula/bottle-feeding compared to non-WIC mothers. WIC and non-WIC mothers’ perceptions of their infants and their feeding styles also differed: WIC mothers reported their infants had significantly lower levels of negative affect, satiety responsiveness, and slowness in eating, and greater enjoyment of food compared to non-WIC mothers and also reported more restrictive and less responsive feeding styles.

Table 1.

Sample Characteristics by WIC Status ^a

	WIC Mothers (n = 46)	Non-WIC Mothers (n = 29)	χ2 or F-value
Infant Characteristics
Sex, % female	54.4 (25)	48.3 (14)	0.26
Age, weeks	14.2 ± 7.0	18.2 ± 7.2	5.72*
Weight-for-length z-score	0.7 ± 1.1	−0.1 ± 1.1	10.47**
Temperament subscales b
Surgency/extraversion	4.3 ± 0.9	4.6 ± 0.6	1.41
Orienting/regulation capacity	5.5 ± 0.9	5.7 ± 0.6	1.06
Negative affect	3.3 ± 1.1	4.2 ± 0.9	10.96**
Eating behavior subscales c
Enjoyment of food	4.5 ± 0.5	4.0 ± 0.5	14.88**
Food responsiveness	2.5 ± 1.0	2.8 ± 0.7	1.60
Satiety responsiveness	2.1 ± 0.6	2.8 ± 0.8	17.16**
Slowness in eating	2.4 ± 0.5	2.9 ± 0.6	16.19**
General appetite	3.4 ± 1.3	3.9 ± 0.7	3.01
Maternal/Familial Characteristics
Age, years	27.4 ± 6.4	32.2 ± 3.9	13.05**
BMI, kg/m2	31.1 ± 6.4	27.0 ± 7.3	6.53*
Parity, % primiparous	37.0 (17)	72.4 (21)	8.95**
Annual Family Income Level			54.57**
<$15,000	41.3 (19)	0.0 (0)
$15,000–$34,999	34.8 (16)	3.4 (1)
$35,000–$74,999	17.4 (8)	10.3 (3)
≥$75,000	0.0 (0)	72.4 (21)
Not reported	6.5 (3)	13.8 (4)
Level of Education			42.23**
Did not complete high school	10.9 (5)	0.0 (0)
High school degree	39.1 (18)	0.0 (0)
Some college	34.8 (16)	13.8 (4)
Bachelors or graduate degree	10.9 (5)	86.2 (25)
Not Reported	4.3 (2)	0.0 (0)
Racial/Ethnic Category			32.00**
Non-Hispanic White	26.0 (12)	75.9 (22)
Non-Hispanic Black	60.9 (28)	3.4 (1))
Hispanic White	2.2 (1)	13.8 (4)
Hispanic Black	6.5 (3)	0.0 (0)
Asian/Pacific Islander	2.2 (1)	6.9 (2)
Native American	2.2 (1)	0.0 (0)
Feeding styles subscales d
Laissez-faire	2.3 ± 0.8	2.6 ± 1.0	2.40
Pressuring	2.2 ± 0.6	2.2 ± 0.5	0.39
Restrictive	3.6 ± 0.8	2.8 ± 0.6	19.92**
Responsive	4.2 ± 0.5	4.5 ± 0.3	10.79**
Feeding Characteristics
Milk Type			36.88**
Exclusively breast milk	8.7 (4)	72.4 (21)
Exclusively formula	76.1 (35)	10.3 (3)
Mix of breast milk and formula	15.2 (7)	17.2 (5)
Feed Mode			35.62**
Exclusively breastfed	2.2 (1)	13.8 (4)
Exclusively bottle-fed	89.1 (41)	20.7 (6)
Mix of breast and bottle	8.7 (4)	65.5 (19)

Note: Values are percent (n) or mean ± SD

^*p<.05;

^**p<.01 for difference between non-WIC and WIC mothers

^aData are from WIC and non-WIC mothers (n=75) with infants under 6 months of age who participated in infant feeding studies

^bFrom the Infant Behavior Questionnaire-Revised Very Short Form;²² possible score range: 1–7

^cFrom the Baby Eating Behavior Questionnaire;²⁴ possible score range: 1–5

^dFrom the Infant Feeding Styles Questionnaire;²³ possible score range: 1–5

Prevalence of maternal distraction during feeding

Table 2 presents results of the thematic analysis of mothers’ feeding records. Mothers reported a variety of additional activities while feeding their infants, which were further categorized into technological versus non-technological distractions. Distractions were reported during 43% of feedings, with mothers reporting technological distractions during 26% of feedings and non-technological distractions during 17% of feedings. For the remaining 57% of feedings, mothers reported interacting with their infants, that they did not do anything else during the feeding, or left the question blank. Reported feeding durations were not associated with reports of any type of distraction (F[1,2784]=0.19, P=.66) or technological distraction (F[1,2784]=0.05, P=0.82) during feeding.

Table 2.

Proportions of feedings where mothers reported distractions versus no distractions ^a

	%	n
Technological Distraction Reported	26.0	775
Watching television	19.5	580
Using a smart phone or tablet	5.4	161
Using a computer	0.9	26
Multiple technologies	0.3	8
Non-Technological Distraction Reported	16.9	504
Talking on the phone or to another adult	4.1	122
Sleeping	4.0	120
Reading a newspaper, book, or magazine	3.1	91
Listening to music	2.0	60
Doing housework/Caring for other children	1.7	52
Eating	1.2	36
Traveling	0.8	23
No Distractions Reported	57.1	1703
Nothing specified	51.8	1544
Interacting with baby	5.3	159

^aData are from WIC and non-WIC mothers (n=75) with infants under 6 months of age who participated in infant feeding studies; these analyses were based on n = 2982 recorded feedings

Of the 2982 reported feedings, 36% (n = 1075) were breastfeedings (i.e., from the breast) and 64% were bottle-feedings (n = 1906; note that feeding mode was not reported for one feeding). Of bottle-feedings, 72% (n = 1367) were formula, 26% (n = 500) were expressed breast milk, and 2% (n = 39) were a combination of breast milk and formula. Table 3 presents the percentages and odds of mothers’ reports of any, technological, or no distractions as a function of feeding mode (breast-versus bottle-feeding) during the recorded feeding. Overall, mothers had similar odds of reporting any distraction (P = .68), technological distractions (P = .37), non-technological distractions (P = .67), or no distractions (P = .68) during breast-versus bottle-feedings.

Table 3.

Odds of reporting technological distractions, non-technological distractions, or no distractions during breast versus bottle-feedings ^a

	Breastfeedings		Bottle-feedings
	%	n	%	n	OR (95% CI)b, c	P-value
Technological Distraction Reported	23.2	249	27.6	526	0.79 (0.48, 1.32)	.37
Watching television	10.7	115	24.3	465
Using a smart phone or tablet	10.3	111	2.6	50
Using a computer	1.5	16	0.5	10
Multiple technologies	0.7	7	0.1	1
Non-Technological Distraction Reported	18.0	193	16.3	311	1.12 (0.66, 1.92)	.67
Sleeping	4.1	44	4.0	76
Talking on the phone or to another adult	3.3	35	4.6	87
Reading a newspaper, book, or magazine	5.5	59	1.7	32
Listening to music	2.9	31	1.5	29
Doing housework/Caring for other children	0.3	3	2.6	49
Eating	1.6	17	1.0	19
Traveling	0.4	4	1.0	19
No Distractions Reported	58.9	633	56.1	1070	1.12 (0.65, 1.93)	.68
Nothing specified	52.5	564	51.4	980
Interacting with baby	6.4	69	4.7	90

^aData are from WIC and non-WIC mothers (n=75) with infants under 6 months of age who participated in infant feeding studies; these analyses were based on n = 2982 recorded feedings

^bFor the logistic regression with estimation via generalized estimating equations model exploring whether feeding mode predicted the probability of mothers’ reports of each activity; breastfeeding was specified as the referent group.

^cWIC participation did not moderate effects of feeding mode on the probability of mothers’ reports of each activity.

Mothers’ intensities of distracted feeding

When analyzed at the level of the individual, 92% of mothers reported a distraction during at least 1 feeding. Eighty-three percent of mothers reported a technological distraction during one or more feedings. Mothers’ reports of distractions ranged from 0% to 100% of recorded feedings (Mean = 43.9%, SD = 29.5%); similarly, reports of technological distractions ranged from 0% to 97% of recorded feedings (Mean = 27.7%, SD = 24.0%).

Associations between mothers’ intensities of distracted feeding and infant and mother characteristics

Table 4 illustrates the best-fit models for prediction of mothers’ intensities of any distraction and technological distractions. Significant predictors of mothers’ intensities of any distraction during feeding were multiparity and mothers’ perception of greater infant appetite. Significant predictors of mothers’ intensities of technological distraction during feeding were mothers’ racial/ethnic minority (e.g., black, Hispanic, or Asian) status, adherence to a more laissez faire feeding style, younger infant age, and perception of lower infant food responsiveness and greater infant appetite. Feeding mode; maternal WIC status, education, and BMI; and infant sex, weight status, and temperament were not significant predictors of distraction in either model.

Table 4.

Associations between characteristics of mothers and infants and mothers’ intensities of any distraction and technological distraction during feeding ^a

Model	Dependent Variable	Independent Variables	Coefficient b	SE	P-Values
1	Intensity of any distraction c (F = 6.38; P = .003; R2 = 0.20)	Multiparous (versus primiparous)	18.27	7.43	.0173
		Infant appetite	10.91	3.47	.0028
2	Intensity of technological distraction d (F = 9.10; P < .0001; R2 = 0.48)	White (versus Black, Hispanic, or other)	−13.56	4.93	.0083
		Mother laissez faire feeding style	6.83	2.75	.0165
		Infant age	−0.83	0.35	.0222
		Infant food responsiveness	−13.66	3.62	<.001
		Infant appetite	11.37	2.64	<.0001

^aData are from WIC and non-WIC mothers (n=75) with infants under 6 months of age who participated in infant feeding studies;

^bFor multiple stepwise regression models exploring the combination of maternal (i.e., feeding mode, education, race/ethnicity, parity, WIC status, BMI, feeding styles) and infant (i.e., sex, age, weight status, temperament, eating behaviors) characteristics that predicted maternal intensity of any or technological distraction.

^cCalculated for each mother as: (number of feedings wherein any distraction were reported)/(total number of feedings reported)*100

^dCalculated for each mother as: (number of feedings wherein a technological distraction were reported)/(total number of feedings reported)*100

DISCUSSION

The present study aimed to describe 1 potential barrier to responsive feeding that is particularly relevant in today’s technology-focused society – the presence of and engagement in distractions during infant feeding. Findings from the present study illustrate that distracted feeding is a common practice for both WIC and non-WIC mothers, regardless of mode of feeding. Although few studies have described the extent to which caregivers engage in distractions during feeding and other caregiver-child interactions, the findings of the present study are consistent with previous studies focused on bottle-feeding dyads¹⁹ and families with older children,^16,17 illustrating that significant proportions of caregivers are frequently engaging with technological and other environmental stimuli during child feeding and mealtime interactions.

The significance of caregiver responsiveness during feeding interactions is underlined by the belief that infants have an innate capacity to self-regulate^35,36 that develops during the first few months postpartum and this development is shaped by the interactional learning that occurs when the caregiver responds contingently and appropriately to infant cues.³⁷ During early infancy, feeding interactions comprise a significant proportion of caregiver-infant interactions; thus, high quality, synchronous feeding interactions provide the infant with both nutritive and socioemotional benefits.^38–43 In light of this understanding, an important question is whether the presence of certain distractions during feeding interactions impacts the responsiveness of caregivers to children’s cues and needs.

The present study tested several hypotheses aimed at understanding whether certain dyads may be more prone to distracted feeding than others. Inconsistent with previous research,¹⁹ hypothesized associations between distracted feeding and breastfeeding; mothers’ age and self-reported tendencies toward responsive feeding; and infants’ temperamental characteristics were not found. However, the present study did support hypothesized associations between distracted feeding and parity and infant eating behaviors. Interpretation of these findings within the broader context of previous research on caregiver feeding practices and styles reveals a number of consistencies. Previous research has similarly found that multiparious mothers report greater levels of distraction during infant feeding,¹⁹ which makes intuitive sense given the increased number of children for which these mothers care. Although further research is needed to verify whether distracted feeding is a facet of non-responsive feeding, the finding that minority race/ethnicity was a significant predictor of technological is consistent with previous research illustrating racial/ethnic minorities are more likely to report using practices characterized by low responsiveness to child cues.^12,44 Additionally, the association between higher levels of technological distraction and adherence to a laissez-faire feeding style makes intuitive sense because laissez-faire feeding style is characterized by feeding practices that are low in involvement and structure (e.g., propping the bottle, watching TV during feeding).³¹ Taken together, these findings may suggest a number of factors, such as family structure, cultural beliefs, and broader feeding practices and styles, influence mothers’ tendencies toward distracted feeding and abilities to engage in responsive feeding.

It is important to note, however, that the limitations of the present study underline possible caveats to this interpretation. First, the correlational and self-report nature of this study precludes abilities to determine cause and effect relationships. For example, because infant eating behaviors (e.g., appetite, food responsiveness) were reported by mothers, it is unclear whether maternal distraction influences the development of infants’ appetite and food responsiveness or whether mothers with younger infants or who perceive their infants to have larger appetites and lower food responsiveness attend to distractions during the feeding for some other reason (e.g., they feel the feeding takes too long or do not know how to engage with their young infant). Second, the main variable of interest was assessed through a qualitative question on an infant feeding diary. A strength of this approach is that it provided rich qualitative data on the wide array of things mothers do while feeding their infants. However, it is also possible that mothers underreported their levels of distraction because self-reported engagement in some distractors tends to be biased, such as mobile devices because bouts of use are short and interspersed through the day.⁴⁵ Relatedly, interpretation of this thematic analysis of maternal reports is also limited by the fact that when a mother left a section blank, it was not known if this was because no distractions were present or if the mother failed to report. Mothers were not provided instruction regarding when to complete the records (e.g. in real time versus at the end of the day); for those mothers who completed the records during the feeding, the records may have actually introduced an additional source of distraction. Finally, although associations between feeding mode and distraction were explored, only 5 of participants were exclusively breastfeeding and 23 were breast- and bottle-feeding; thus, although a substantial number of breastfeedings were recorded, these feedings came from a limited number of participants, which limits the generalizability of the findings.

IMPLICATIONS FOR RESEARCH AND PRACTICE

An important focus for future research is to determine what impact, if any, distracted feeding has on the quality and outcome of the feeding interaction in the short-term, and on infants’ developing eating behaviors and self-regulatory abilities in the long-term. Additionally, given evidence for inter-individual variability in mothers’ level of responsiveness during feeding interactions,^{37,43,46–49} future research should also explore whether certain psychosocial factors, such as responsive feeding skills, postpartum depression, participation in WIC or other federal assistance programs, or cultural beliefs, moderate the potential impact of distractors on feeding outcomes. This research would elucidate whether distracted feeding is a form of non-responsive feeding, or whether some mothers are still able to feed responsively, even in the presence of technological or other distractions.

Further research using experimental designs and/or observational measures of maternal and infant behavior in the presence of environmental distractors is needed to better understand the extent to which mothers attended to distractors versus their infant during feeding and the extent to which the presence of distractors affects feeding interactions. If future research suggests maternal distraction decreases the quality of the feeding interaction, targeted efforts within both research and practice should focus on educating mothers and other caregivers about the potential effects of caregiver distraction on infant feeding and developmental outcomes. Efforts to develop and evaluate approaches to help caregivers better understand and attend to their infants’ cues, especially in the presence of technological and other distractors, may also be warranted.