Integrative Review of Mental Health and Feeding Styles in Parents of Bottle-Fed Infants

Abstract

Objective:

To review the literature regarding the relationship between symptoms of anxiety and depression and feeding styles in parents of bottle-fed infants.

Data Sources:

We conducted literature searches in PubMed, CINAHL, Scopus, and PsycINFO.

Study Selection:

We used Covidence systematic review management software during the selection process to allow for full blinding of decisions by team members. Articles were eligible for inclusion if they were reports of primary research, written in English, and focused on the relationship between symptoms of anxiety or depression and feeding styles in parents of term, bottle-fed infants less than 12 months of age. We placed no restriction on date of publication because of the sparse amount of published literature on this topic. We identified a total of 1,882 articles. After removing duplicates, we screened 988 articles and retained six articles that met criteria for our review.

Data Extraction:

We used Whittemore and Knafl’s integrative review methodology to guide data extraction and reporting. We extracted relevant data from all primary data sources and compiled the data into a matrix. We used The Joanna Briggs Institute Checklist for Analytical Cross-Sectional Studies to assess the quality of the studies.

Data Synthesis:

Using an a priori coding scheme, we summarized the data using categorization of established parental feeding styles in infancy applicable to bottle-feeding. We synthesized the data into two broad categories: responsive and non-responsive feeding styles.

Conclusion:

Although research on the topic is limited, our findings suggest that symptoms of postpartum depression may be associated with non-responsive feeding styles in parents of bottle-fed infants. We suggest several areas for future research and recommend increased emotional and feeding support in practice for parents of bottle-feeding infants.

Precis

Symptoms of depression may be associated with non-responsive feeding styles in parents of bottle-fed infants and indicate a need for increased emotional and feeding support.

Callouts

1. Non-responsive feeding styles have been associated with rapid weight gain in infancy, and formula-fed infants more commonly experience rapid weight gain than breastfed infants.

2. Symptoms of depression may be associated with non-responsive feeding styles in parents of bottle-fed infants.

3. There is a need for health care professionals to provide emotional support and promote responsive feeding in parents of bottle-fed infants.

While breastfeeding has been shown to decrease risk of childhood obesity (Anderson et al., 2020; Yan et al., 2014), many families choose not to or are unable to breastfeed for various reasons (Abrams & Daniels, 2019). In fact, 75% of infants in the United States receive some form of non-breastmilk food as a primary source of nutrition during the first 6 months of life (Centers for Disease Control and Prevention [CDC], 2020). Not only the milk type, but also the parent’s style of engagement during infant feeding sessions can influence an infant’s risk for obesity (Ventura et al., 2022). Parental feeding styles encompass the overall attitudes, behavior, and emotional climate a parent creates with the infant during feeding sessions (Thompson et al., 2009; El-Behadli et al., 2015). There are two main types of feeding styles: responsive and nonresponsive. The American Academy of Pediatrics (2017) recommended responsive feeding, being attentive to the infant’s hunger and satiety cues, which has been shown to facilitate healthier weight trajectories in children less than 2 years (Spill et al., 2019). Non-responsive feeding styles encompass all types of feeding styles in which parents are not attuned or attentive to infant hunger and satiety cues (Table 1; Pérez-Escamilla et al., 2017). Non-responsive feeding styles may disrupt appetite self-regulation by compromising an infant’s ability to recognize internal satiety cues (Hodges et al., 2020), and non-responsive feeding styles have been associated with rapid weight gain in infancy (Appleton, Russell, et al., 2018), which increases risk for childhood obesity (Zheng et al., 2018).

Table 1.

Feeding Styles in Infancy Applicable to Bottle-Feeding

Responsive	Non-Responsive
Responsive	Pressuring
Definition: “The parent is attentive to child hunger and satiety cues” (Thompson et al., 2009, p. 211). Associated Practices: No distraction during feeds Attentive to hunger and satiety cues Feeding “on demand”	Definition: “The parent is concerned with increasing the amount of food the infant consumes and uses food to soothe the infant” (Thompson et al., 2009, p. 211). Associated Practices: Adding cereal to bottle Feeding infant to make them stop crying Feeding infant in the absence of hunger cues/ feeding on strict schedule Continuing to feed when satiety cues are exhibited by infant
	Restrictive
	Definition: “The parent limits the infant to healthful foods and limits the quantity of food consumed” (Thompson et al., 2009, p. 211) Associated Practices: Limiting amount of intake Not feeding infant when hunger cues are exhibited/ feeding on strict schedule
	Laissez-faire
	Definition: “The parent does not limit infant diet quality or quantity and shows little interaction with the infant during feeding” (Thompson et al., 2009, p. 211). Associated Practices: Distractions (i.e., Watching television while feeding infant) Propping the bottle

Parental mental health symptoms have been linked to non-responsive feeding styles (El-Behadli et al., 2015), and rates of postpartum anxiety and depression are concerning globally. In a meta-analysis of 102 studies conducted across 34 countries, Dennis et al. (2017) found a prevalence of 15% for maternal symptoms of anxiety postnatally, and in a systematic review of 43 studies conducted across 12 countries, Leach et al. (2016) estimated a prevalence of 2–18% for anxiety disorders among fathers of infants. Additionally, in a meta-analysis of 291 studies across 56 countries, Hahn-Holbrook et al. (2018) found a prevalence of 18% for postpartum depression among women, and in a meta-analysis of 47 studies across 21 countries, Rao et al. (2020) estimated a prevalence of 9% for paternal postpartum depression. Contributors to postpartum anxiety and depression include hormonal changes and sleep disruption (Zappas et al., 2021), social identity struggles (Seymour-Smith et al., 2017), and lack of financial resources (Atkins et al., 2020), and mothers who formula-feed may experience greater symptoms of anxiety and depression than those who breastfeed (Penniston et al., 2021) possibly placing them at greater risk for engaging in non-responsive feeding styles. There is a need to review the literature related to potential contributors to non-responsive feeding styles, such as symptoms of anxiety and depression in parents, to identify those at high-risk and intervene appropriately to encourage responsive feeding.

Callout 1

Feeding styles apply to breast and bottle-feeding, but given the link between formula feeding and increased risk for childhood obesity (Watchmaker et al., 2020; Whaley et al., 2017) and parental mental health symptoms (Penniston et al., 2021), our initial aim was to synthesize reports of studies that included infants who were strictly formula-fed. Upon reviewing the literature, however, we found that many researchers did not specify what nutritional substance (i.e., expressed breastmilk or formula) was given to the infant. As a result, we broadened the scope of the review to include bottle-fed infants. Thus, the purpose of this integrative review was to review the literature regarding the relationship between symptoms of anxiety and depression and feeding styles in parents of bottle-fed infants.

Methods

We used Whittemore and Knafl’s (2005) methodology for our integrative review. This framework includes five stages: problem identification, literature search, data evaluation, data analysis, and presentation. After we identified the problem, we conducted a literature search to explore the relationship between symptoms of anxiety and depression and feeding styles among parents of bottle-fed infants.

Literature Search

In collaboration with a health sciences research librarian, we conducted searches in PubMed, CINAHL, Scopus, and PsycINFO using the following keywords: depression, depressive, baby blues, anxiety, mental health, psychosocial, parents, parental, mothers, maternal, fathers, caregivers, guardians, beliefs, practices, attitudes, behavior, styles, factors, feeding, formula, bottle feeding, infant, infancy, neonate, and newborn with various combinations and using Boolean operators. We included articles if they were reports of primary research studies published in English in which researchers examined symptoms of anxiety or depression and their associations with parents’ feeding styles of term, bottle-fed infants less than 12 months of age. We excluded articles if they had samples that did not include parents of infants less than 12 months of age, had samples primarily of parents of preterm infants, were focused on breastfeeding, or were not peer-reviewed reports of primary research. We placed no restriction on date of publication. We conducted the original literature search in February 2021 and updated it on September 2, 2022.

We used Covidence systematic review management software (Veritas Health Innovation Ltd., 2021) throughout the review processes. This software increases the rigor of the review because it allows for full blinding of decisions by team members during the screening process and it prompts users to engage in the review process in a systematic order (Kellermeyer et al., 2018). We have presented the results of the screening and selection process in a PRISMA diagram (Page et al., 2021) in Figure 1. Our initial search resulted in 1,882 articles. After removing duplicates, we screened 988 articles by title and abstract. Two reviewers (TN and CC) independently screened articles by title and abstract, and one reviewer (DT) resolved conflicts as needed. After we screened by title and abstract, 169 articles remained for full-text review. For this phase, TN reviewed all articles, and CC and DT each reviewed half of the articles; CC and DT resolved conflicts for full-text articles they did not originally review. Of the full-text articles screened, we excluded 135 because feeding styles were not examined as an outcome variable (n = 70), the article was not peer-reviewed primary research (n = 30), or parents’ symptoms of anxiety or depression were not study variables (n = 14), among other reasons (see Figure 1). After we eliminated these articles, 34 articles remained for further evaluation.

Figure 1.

PRISMA diagram of search strategy (Page et al., 2021).

Data Evaluation

At least two authors independently reviewed all articles. Following the independent review, we collaboratively participated in a 2-day intensive analysis retreat. This dedicated time allowed for full immersion in the data and was pivotal for the data evaluation and analysis stages of the integrative review (Whittemore & Knafl, 2005). During the retreat, we engaged in quality appraisal and selection of articles to minimize researcher bias. We collaboratively evaluated the remaining 34 articles and found that many authors did not stratify the results by feeding method (i.e., breastfeeding v. bottle-feeding). Because our aim was to review the literature regarding the relationship between symptoms of anxiety and depression and feeding styles in parents of bottle-fed infants, we eliminated studies in which authors did not stratify results by feeding method (n = 28). Six articles remained for further analysis (see Figure 1). All studies had descriptive, correlational quantitative designs and were either cross-sectional or the data were drawn from one time point in a longitudinal study. Whittemore and Knafl (2005) noted that when primary sources have similar designs, incorporating quality appraisal scores into the analysis may be optimal. Thus, we used the Joanna Briggs Institute (2017) Checklist for Analytical Cross-Sectional Studies for quality appraisal.

After independently appraising the studies, we met to discuss and resolve any discrepancies. In all six studies, researchers measured the independent variables (symptoms of anxiety or depression) using tools with evidence of validity and reliability. However, only Savage and Birch (2017) used a validated tool to measure the outcome variable (feeding styles), and Lucas et al. (2017) used only two of the 28 questions from the Infant Feeding Questionnaire (Baughcum et al., 2001). In a secondary data analysis by Wasser et al. (2011), mothers completed an infant diet history and 24-hour diet recall during home visits in the parent study to assess early introduction of solid foods, including cereal in the infant’s bottle. The parent study also included additional 24-hour diet recalls on two random dates by telephone within 2 weeks of the home visit to increase accurate descriptions of the infant’s typical dietary intake. In the remaining three studies (Hellin & Waller, 1992; Gaffney et al., 2014; Paulson et al., 2006), authors did not use tools with evidence of validity or reliability for measuring feeding styles. In all but one study (Hellin & Waller, 1992), authors clearly defined inclusion criteria for the sample, described the subjects and setting in detail, and discussed using strategies to deal with confounding factors. As our purpose was to examine the literature on this topic, we included all articles in the analysis regardless of methodological rigor.

Data Analysis

In the data analysis stage of Whittemore and Knafl’s (2005) integrative review methodology, data are coded, categorized, and then synthesized to form conclusions from the findings. This stage includes five processes: data reduction, data display, data comparison, conclusion drawing, and verification. Our author group brings integrated expertise consistent with the aim of this review including mental health (DT) and maternal-newborn nursing (TN & CC), which enriched the analysis. We first analyzed the six articles independently, then collaboratively. We used an a priori coding scheme; we reduced the data by categorizing the findings according to the established parental feeding styles in infancy applicable to bottle-feeding (see Table 1) as outlined by Thompson and colleagues (2009). Some researchers discussed feeding practices rather than feeding styles, such as putting an infant to bed with a bottle or bottle-feeding on strict time intervals, which could correspond to specific feeding styles. Therefore, we coded the data using the broad categories of responsive and non-responsive feeding styles to be consistent with the published terminology and to avoid interpretation of feeding styles not explicitly indicated within the articles. As an analytic exercise, we displayed the data with visuals created by each author depicting patterns and relationships between parental mental health and feeding styles. We discussed our visual representations to facilitate consensus of our findings. This exercise permitted our team to conduct deeper analysis, conclusion drawing, and verification of the data. Finally, we summarized and presented the data in a uniform matrix (see Table 2).

Table 2.

Summary of Studies Depicting the Relationship Between Depression or Anxiety and Feeding Styles in Infancy

Author/Date	Aim/Purpose	Participants	Methods	Outcomes	Feeding Style(s)
Hellin & Waller (1992)	“Explore the associations between mothers’ mood, their perception of infant feeding difficulties and their feeding practices.”	N= 145 mothers Mean age: 25.7 years 62.5% multiparous 58.5% BF at least once Average infant age: 5 months	Prospective Longitudinal Descriptive Correlational Collected data @ 3 time points over 1 year (during pregnancy, within 1 week of delivery, & 17–27 weeks postpartum). Data related to feeding practices was only collected at final time point. Anxiety sx: STAI Depressive sx: BDI Feeding Practices: researcher-developed questionnaire	Feeding Practices: 29.7% scheduled 27.7% on demand 42.5% both routines Anxiety and depressive sx did not predict whether women fed on demand, scheduled, both	Responsive (“on demand”) Non-responsive (scheduled)
Paulson et al. (2006)	“Examine the effects of maternal and paternal depression on parenting behaviors consistent with anticipatory guidance recommendations.”	N=5,089 2-parent families Sample from 9-month-old wave of the national ECLS by NCES 74.4% children ever BF 51.6% White 14% mothers & 10% fathers with moderate or severe depressive sx Average infant age: 9 months	Secondary Data Analysis From 9-month data collection point of ECLS Depressive sx: CES-D Feeding practice (bed with bottle): maternal interview dichotomous yes/no question	Mothers with depressive sx more likely to put infant to bed with bottle (OR = 1.53, 95% CI [1.16, 2.02]; p < .01) When both parents had depressive sx, infant more likely to be put to bed with a bottle (X2 [3] = 45.74; p < .001) Adjusted effect: (OR = 1.08, 95% CI [0.54, 2.16])	Non-responsive
Wasser et al. (2011)	“Assess early infant-feeding patterns in a cohort of low-income Black mothers and to examine associations between maternal perception of infant temperament and CF before 4 months.”	N=217 mother-infant dyads Sample from ICFROS, which was recruited through NC WIC Primiparous 29% depressive symptoms African American Term, healthy infants BF @ 3 months: 25% some BF, 5% EBF Average infant age: 3 months	Secondary Data Analysis Data from original study was collected through home visits at infants ages 3, 6, 9, 12, & 18 months (data from this study reflective of 3-month visit) Data collected from 2003–2007 Depressive sx: CES-D Feeding practice: infant diet history, 24 hour diet recall	Infant cereal in the bottle was the most commonly fed solid food Cereal in Bottle Rates 1 month: 6% 2 months: 41.5% 3 months: 54.3% Mothers with depressive sx more commonly introduced solids early compared to mothers without depressive sx mothers (p > .05)	Non-responsive pressuring
Gaffney et al. (2014)	“Examine the effects of PPD on early infant feeding practices, including BF intensity at 2 months of age, adding baby cereal to formula at 2 months of age, and early intro solids (< 4 months of age).”	N=1,447 mother-infant dyads Sample from IFPSII Healthy infants 24.1% mothers with depressive symptoms 91% White Average infant age: 2 months (cereal in bottle)	Secondary Data Analysis Parent Study: Longitudinal study- mailed surveys (×1 prenatal & ×10 postnatal) mailed monthly Depressive sx: 10-item EPDS Feeding practice (cereal in bottle): dichotomous yes/no survey question	Mothers with depressive sx more likely to add cereal to bottle (OR = 1.77, 95% CI [1.16, 2.68])	Non-responsive pressuring
Lucas et al. (2017)	“Determine maternal and infant characteristics associated with adding cereal into the bottle.”	N=216 mother-infant dyads Sample from BFELLE 88% low SES 91% Latina 10% mothers with moderate or severe depressive sx Term, healthy infants not EBF Average infant age: 6.5 months	Secondary Data Analysis Parent Study (longitudinal): Interviews for read-aloud surveys in Spanish/English Cereal in bottle data @ baseline & infant age 6 months & depressive sx data @ infant age 6 months for this analysis Depressive sx: PHQ-9 Feeding practice (cereal in bottle): two questions from Infant Feeding Questionnaire	Moderate to severe depressive sx associated with adding cereal to bottle (OR = 4.1, 95% CI [1.4, 12.2]; p = .01)	Non-responsive pressuring
Savage & Birch (2017)	“Explore how maternal depressive symptomology is related childhood obesity-promoting parenting behaviors, and if depressive symptomology moderates the association between perceived child negativity and the use of food to soothe among low-income mothers.”	N=60 mothers + formula-fed infants/toddlers (0–2 years old) WIC-enrolled 75% White, 19% African American 46% annual income <15K/year 38% with elevated depressive sx Average infant age: 11.5 months	Cross-sectional Descriptive Correlational Recruited through WIC clinics Depressive sx: CES-D Feeding Practices: Baby’s Basic Needs Questionnaire	Mothers with depressive sx reported using food to soothe more often than mothers without depressive sx (p < .05) Mothers with depressive sx more likely to use food to soothe in various situations regardless of perceived infant fussiness (R2 = 0.56, p < .05) 83% of mothers with depressive sx put cereal in bottle compared to 58% mothers without depressive sx 70% of mothers with depressive sx put infant to bed with bottle compared to 33% mothers without depressive sx (p < .01)	Non-responsive

Note. ECLS: Early Childhood Longitudinal Study; NCES: National Center for Education Statistics; ICFROS: Infant Care, Feeding and Risk of Obesity Study; IFPSII: Infant Feeding Practices Study II; BFELLE: Bellevue Project for Early Language, Literacy and Education Success RCT; SES: Socioeconomic Status; WIC: Supplemental Nutrition Program for Women, Infants, and Children; BF: Breastfeeding; EBF: Exclusive Breastfeeding; CF: Complementary Feeding; PPD: Postpartum Depression STAI: State-Trait Anxiety Inventory; BDI: Beck Depression Inventory; CES-D: Center for Epidemiological Disease Depression; EPDS: Edinburgh Postnatal Depression Scale; PHQ-9: Patient Health Questionnaire Depression Module; OR: Odds Ratio; CI: Confidence Interval; Sx: Symptoms

Results

We included six articles in our review which were published from 1992 to 2017. One study was conducted in the United Kingdom (Hellin & Waller, 1992) and five studies were conducted in the United States (Gaffney et al., 2014; Lucas et al., 2017; Paulson et al., 2006; Savage et al., 2017; Wasser et al., 2011). All studies had descriptive, correlational quantitative designs with data from one time point, and four were secondary data analyses (Gaffney et al., 2014; Lucas et al., 2017; Paulson et al., 2006; Wasser et al., 2011). Sample sizes of these studies varied greatly from 60 (Savage & Birch, 2017) to 5,089 participants (Paulson et al., 2006). Only Savage and Birch (2017) included a full sample of exclusively formula-fed infants. In the other five studies (Gaffney et al., 2014; Hellin & Waller, 1992; Paulson et al., 2006; Lucas et al., 2017; Wasser et al., 2011), researchers included participants who were breastfeeding or bottle-feeding. However, they specifically reported findings related to symptoms of anxiety or depression on bottle-feeding practices (Gaffney et al., 2014; Paulson et al., 2006; Lucas et al., 2017; Wasser et al., 2011) or that there were no significant differences in mood between breast and bottle-feeding participants and mood did not predict whether participants fed their infants on a fixed schedule or when their infants exhibited hunger cues (Hellin & Waller, 1992).

Hellin and Waller (1992) were the only researchers to measure symptoms of anxiety, and they used the State-Trait Anxiety Inventory (STAI). All researchers measured symptoms of depression using standardized instruments, including the Beck Depression inventory (BDI; Hellin & Waller, 1992), the ten-item Edinburgh Postnatal Depression Scale (EPDS; Gaffney et al., 2014), the nine-item Patient Health Questionnaire Depression Module (PHQ-9; Lucas et al., 2017), and the Center for Epidemiologic Studies Depression Scale (CES-D; Paulson et al., 2006; Savage & Birch, 2017; Wasser et al., 2011). While authors of three studies used the CES-D to measure symptoms of depression, cutoff scores for depression differed for one study (Paulson et al., 2006). Further details of each article are listed in Table 2. We synthesized the articles in this review to best represent how symptoms of anxiety and depression were related to responsive and non-responsive feeding styles.

Callout 2

Responsive Feeding Style

Hellin and Waller (1992) reported on the relationships between symptoms of anxiety and depression and the responsive feeding style. In their prospective, longitudinal descriptive correlational study, Hellin and Waller (1992) distributed a questionnaire to 145 mothers of infants to obtain their perceptions of feeding difficulties and practices. Specifically, the researchers captured feeding styles by asking mothers if they initiated feedings when their infants exhibited hunger cues (demand), at predetermined times (scheduled), or both. For the 101 women who responded to this question, 28% fed on demand, 30% fed at scheduled times, and 43% used both practices. For participants who exhibited a responsive feeding style and fed their infants on demand, symptoms of anxiety and depression did not predict their feeding practices.

Non-Responsive Feeding Styles

Non-responsive feeding styles (Table 1) encompass all types of feeding styles that deviate from the optimal responsive feeding style. Parents who exhibit non-responsive feeding styles are not attuned or attentive to infant hunger and satiety cues (Pérez-Escamilla et al., 2017). Hellin and Waller (1992) found that symptoms of anxiety and depression did not predict non-responsive feeding styles, including the practice of feeding on fixed time intervals (Hellin & Waller, 1992). Researchers of two studies found that parents who experienced more symptoms of depression were more likely to put their infants to bed with a bottle (Paulson et al., 2006; Savage & Birch, 2017). Of note, Paulson et al. (2006) were the only researchers to include fathers in their study sample. They found that when both parents had symptoms of depression, the infant was more likely to be put to bed with a bottle (X² [3] = 45.74; p < .001); however, after adjusting for covariates, the logistic regression analysis did not reveal a statistically significant relationship (OR = 1.08, 95% CI [0.54–2.16]) between both parents having symptoms of depression and the infant being put to bed with a bottle. Researchers of three studies found that mothers with more symptoms of depression were more likely to put cereal in their infants’ bottles (Gaffney et al., 2014; Lucas et al., 2017; Savage & Birch, 2017), which is specifically consistent with the pressuring feeding style (Thompson et al., 2009). Similarly, Wasser et al. (2011) found that mothers with symptoms of depression more frequently introduced solid foods early. The most common solid food was cereal that was added to the bottle, but this relationship was not statistically significant (Wasser et al., 2011). Savage and Birch (2017) also found that mothers with symptoms of depression exhibited the pressuring feeding style through the practice of using food to soothe their infants (Thompson et al., 2009). Of note, Savage & Birch (2017) were the only researchers who used a validated subscale to measure feeding styles. In the Baby’s Basic Needs Questionnaire Food to Soothe subscale items, mothers were asked, “How often did you offer food or beverage to soothe your child?” and how likely were they to use food to soothe in various settings and situations (Savage & Birch, 2017).

Discussion

Although we aimed to review the literature regarding the relationship between symptoms of anxiety and depression and feeding styles in parents of bottle-fed infants, our findings are inconclusive. Regarding symptoms of anxiety, we found only one study in which researchers examined their relationship with parent feeding styles (Hellin & Waller, 1992). Researchers in all six studies examined the relationship between feeding styles and symptoms of depression in parents of infants, and findings suggest symptoms of depression may be associated with non-responsive feeding styles. However, the use of multiple instruments and measurement strategies hindered our ability to make overall conclusions. Findings that suggest symptoms of depression may be associated with non-responsive feeding styles are consistent with results from other studies on the relationships between symptoms of depression and feeding styles in parents of infants (Chen et al., 2020; Gaffney et al., 2018; Joshi & Raut, 2019), though we were not able to include these articles in our review because findings were not stratified by feeding method (i.e., breastfeeding v. bottle-feeding). Non-responsive feeding is of concern during infancy because of the associated health implications. For example, non-responsive feeding may interfere with the infant’s ability to self-regulate appetite (Hodges et al., 2020), and researchers have found a link between the pressuring feeding style and increased weight gain in infancy (Appleton, Russell, et al., 2018; Stifter & Moding, 2015; Thompson et al., 2021).

Infants who experience rapid weight gain are at greater risk for becoming overweight or obese across the lifespan (Appleton, Russell, et al., 2018; Zheng et al., 2018). What and how infants are fed can influence their risk for rapid weight gain in infancy and later childhood obesity (Ventura, 2022). In a sample of 8,030 infants in the United States, Gibbs and Forste (2013) found that predominantly formula-fed infants were 2.5 times as likely to be obese at 24 months than predominantly breastfed infants. Other researchers found that exclusively formula-fed infants were more likely to be obese in their early childhood years than breastfed infants (Watchmaker et al., 2020; Whaley et al., 2017). Further, formula-fed infants more commonly experience rapid weight gain than breastfed infants (Appleton, Russell, et al., 2018), which may increase their odds up to 3.66 times for being overweight or obese in childhood and adulthood (Zheng et al., 2018). While bottle-feeding can permit parents to have more control over infants’ feeds compared to feeding directly from the breast, potentially contributing to non-responsive feeding styles (Ventura, 2022), engagement in responsive feeding facilitates healthier growth trajectories in infants (Spill et al., 2019). These findings from previous studies combined with our findings in this review highlight the need to provide responsive feeding support for parents who bottle-feed their infants, particularly those experiencing symptoms of depression.

Of the six studies, only Paulson et al. (2006) included fathers in their sample. However, the researchers included the item assessing the behavior of putting the infant to bed with a bottle in the mother’s interview. Therefore, we cannot determine if this non-responsive practice of putting the infant to bed with a bottle was performed by the mother or father. In most of the published literature related to paternal involvement in infant feeding, researchers have focused on fathers’ roles in decision-making related to their infants consuming breastmilk or formula (Earle & Hadley, 2018) or supporting their partners in breastfeeding (Koksal et al., 2022; Ngoenthong et al., 2020). Many families choose to bottle-feed so caregivers other than the mother can assist with infant feeds (Gallo et al., 2019), yet father-infant feeding interactions are less studied. Fathers may experience symptoms of anxiety and depression during the postpartum period (Leach et al., 2016; Rao et al., 2020), and our findings suggest that symptoms of depression may be associated with nonresponsive feeding styles. Therefore, in future studies, researchers addressing mental health and bottle-feeding in infancy should include fathers to understand their unique needs related to emotional well-being and engagement in infant feeding interactions.

One of the challenges of our review was that the age range of infants in the samples varied across developmental stages of infancy. For example, Savage and Birch (2017) reported on the use of food to soothe by mothers with symptoms of depression, but the average age of infants in this study was 11.5 months. The American Academy of Pediatrics recommends introduction of complementary foods around 6 months of age if the child is developmentally ready (Meek et al., 2022). Therefore, given that the average age of the children was older than 6 months in this study, it is unclear if the mothers used solid foods or bottle-feeding to soothe their infants. While researchers in previous studies have demonstrated associations between the use of food to soothe and increased weight status in infants and children, the samples have similarly included infants older than six months who have been introduced to complementary foods (Jansen et al., 2019; Stifter & Moding, 2015). In a recent study, Hupp et al. (2022) sought to address this gap of understanding the use of food to soothe during the period of exclusive milk feeding and found that in mothers of infants ages birth to six months, the use of food to soothe was associated with the pressuring feeding style and increased infant fussiness but not infant weight for age. Using food to soothe can begin in early infancy; thus, more research is needed to examine the behavior of using bottle feeding to soothe and the potential influences of parental mental health symptoms during the first six months postpartum.

The recent COVID-19 pandemic has exacerbated the mental health of adults (Vahratian et al., 2021), and it has particularly had a significant effect on postpartum women. For example, in a sample of 670 postpartum women residing in the United States, over 30% of participants screened positive for symptoms of postpartum depression during the COVID-19 pandemic (Shuman et al., 2022), which exceeds the global prevalence rate of 18% (Hahn-Holbrook et al., 2018). In this study, mothers who formula-fed and mothers who were worried about themselves or their infants contracting the COVID-19 virus were more likely to screen positive for symptoms of postpartum depression (Shuman et al., 2022). As none of the studies included in our review were conducted during the COVID-19 pandemic, researchers should examine the potential effects of the pandemic on symptoms of anxiety and depression in parents of bottle-fed infants and their relationships with feeding styles in future studies.

Another concern is the stigma associated with formula-feeding (Moss-Racusin et al., 2020). Formula-feeding mothers in previous studies have reported feeling guilty and shamed (Fahlquist, 2016; Hvatum & Glavin, 2016; Jackson et al., 2021; Lagan et al., 2014; Thomson et al., 2015), isolated (Thomson et al., 2015), and inadequate as mothers (Fahlquist, 2016; Jackson et al., 2021; Thomson et al., 2015). These results combined with our finding that suggests symptoms of depression may be associated with non-responsive feeding styles in parents of bottle-fed infants raise concern for the mental health of parents. Stigma has been associated with mental health disorders, including depression and anxiety, in other populations including people living with HIV (Armoon et al., 2022) and obesity (Emmer et al., 2020); therefore, it is conceivable that stigma associated with bottle-feeding could negatively affect mental health in parents of infants and we suggest researchers address this gap in future studies.

Formula-feeding mothers also report receiving inadequate information and conflicting advice from health care professionals about how to feed their infants (Appleton, Laws, et al. 2018; Dattilo et al., 2020; Fahlquist, 2016; Fallon et al., 2016; Hvatum & Glavin, 2016; Jackson et al., 2021; Lagan et al., 2014; Thomson et al., 2015). Health care providers have an obligation to adhere to the ethical principle of autonomy, which refers to individuals’ rights to make decisions related to their care and the care of their children and health care professionals’ expectations to respect those choices (King, 2017). In the context of infant feeding, it is part of health care providers’ professional ethos to provide information for parents to make informed choices related to breastfeeding and formula-feeding and to respect their decisions. Building trust with parents and intentional parent-provider communication has positively predicted adherence to guidelines such as infant safe sleep and newborn Vitamin K injections (Fuzzell et al., 2018). Likewise, fostering parent-provider relationships has the potential to improve adherence to infant feeding recommendations.

In 1991, The Baby Friendly Hospital Initiative (BFHI) was launched by the World Health Organization (WHO) and United Nations Children’s Fund (UNICEF; 2018) to encourage breastfeeding support in maternity care centers. The initiative’s “10 steps to successful breastfeeding” has been successful in increasing breastfeeding rates, and continuation of this support for breastfeeding families is certainly warranted (Pérez-Escamilla et al., 2016). That said, the revised Step 6 states that “mothers who are feeding breastmilk substitutes, by necessity or by choice, must be taught about safe preparation and storage of formula and how to respond adequately to their child’s feeding cues” (WHO & UNICEF, 2018, p. 25). This revision includes helping mothers who choose to formula-feed to do so safely and responsively. Additionally, we recognize we cannot assume a bottle-fed infant is exclusively consuming formula. Over the past decade, there has been a rise in mothers’ choices to exclusively pump and provide breastmilk to their infants via bottle; this method has been a favorable choice for mothers who desire their infants to receive human milk but do not want to or are unable to directly breastfeed (Rosenbaum, 2022). Reasons for feeding decisions are multifactorial (Moran-Lev et al., 2021; Moss et al., 2021; Roll & Cheater, 2016), and lack of emotional support for bottle-feeding parents may contribute to mental health issues, non-responsive feeding styles, and negative health outcomes. We encourage health care professionals to promote safe and responsive bottle-feeding for parents who use this infant feeding method. Lastly, as thousands of families in the United States were affected by the acute infant formula shortage of 2022 (Abrams & Duggan, 2022), there is a critical need to understand how this crisis affected their mental health and feeding styles and how we can better support parents who formula-feed.

Callout 3

Limitations

In consultation with a health sciences research librarian, we selected the most relevant databases; however, our search was limited to four databases and did not include ancestry searching or searches of grey literature. Therefore, it is possible that we did not include some published literature on this topic. Additionally, we originally set out to examine articles related to symptoms of anxiety and depression and feeding styles in parents of formula-fed infants; however, we modified the focus to bottle-feeding because of the lack of specificity related to nutritional substances (breastmilk v. formula) in bottles in the current published literature. We recognize that broadening the scope to bottle-feeding is a limitation on delineating the effects of feeding method (bottle v. breast) versus nutritional substance.

Implications

Our findings highlight several areas for future research. We found only one study examining the relationship between symptoms of anxiety and feeding styles. Since researchers have detected a prevalence of postpartum anxiety among men and women around the world (Dennis et al., 2017; Leach et al., 2016), additional research is needed to determine if symptoms of anxiety influence feeding styles in parents who bottle-feed their infants. Additionally, researchers of only one study included fathers in the sample (Paulson et al., 2006) and none included gender-diverse parents; future researchers should seek to understand the effects of mental health on feeding styles in these underrepresented groups. While most studies we included were secondary data analyses examining specific practices associated with feeding styles such as adding cereal to the bottle (Gaffney et al., 2014; Lucas et al., 2017; Wasser et al., 2011) or putting infants to bed with a bottle (Paulson et al., 2006), only Savage and Birch (2017) used a validated subscale to measure feeding styles. In future studies, we recommend the use of a validated responsive feeding instrument for infants less than one year (Heller & Mobley, 2019). Since parental self-report was commonly used to measure feeding styles, researchers should examine the effects of mental health on feeding styles by incorporating observational methods in future designs. Notably, we excluded 28 articles from this review due to study results not being stratified by feeding method (i.e., breastfeeding v. bottle-feeding). While researchers have not previously felt the need to stratify results, there is a critical need to do so in future studies due to the known link between formula-feeding and obesity (Appleton, Russell, et al., 2018; Gibbs & Forste, 2013; Zheng et al., 2018).

Conclusion

While research is limited, our synthesis of the findings of the articles included in our review suggests symptoms of depression in parents of bottle-fed infants may be associated with non-responsive feeding styles. Additionally, we suggest several areas for future research such as the inclusion of fathers, gender-diverse parents, and other caregivers who bottle-feed; the examination of the relationships between symptoms of anxiety and feeding styles; and use of validated tools and observational methods to measure feeding styles. Ultimately, these findings will be useful in informing interventions aimed at optimizing parental well-being to promote responsive feeding in infancy and subsequently mitigate childhood obesity.

Biographies

Taylor Nelson, PhD, RN, is a post-doc trainee, School of Nursing, UNC-Chapel Hill, Chapel Hill, NC.

Courtney Caiola, PhD, MPH, RN, CNE, is an assistant professor, College of Nursing, East Carolina University, Greenville, NC.

Deborah Tyndall, PhD, RN, CNE, is an associate professor, College of Nursing, East Carolina University, Greenville, NC.