Maternal Psychological Factors and Controlled Child Feeding Practices in Relation to Child Body Mass Index

Abstract

Background:

The rise of childhood obesity in Western society has focused attention on parental feeding practices. Despite evidence that controlled feeding influences child weight, there is a paucity of research examining predictors of controlled feeding. The aim of this study was to determine whether maternal antenatal and/or concurrent anxiety and depressive symptoms, including stress, predicted controlled feeding and whether maternal controlled feeding practices, in turn, predict child BMI.

Methods:

In total, 203 mothers participated in a longitudinal follow-up survey. Mothers' self-reported symptoms of anxiety and depression were measured both in pregnancy and at 2–7 years postpartum. Maternal-reported child BMI and maternal use of restriction, pressure to eat, and monitoring were measured at 2–7 years postpartum.

Results:

Feeding practices were not uniformly predictive of child BMI. Maternal use of restriction and monitoring were partially positively predicted by concurrent maternal stress and negatively partially predicted by concurrent depression. Thus, mothers enduring high stress appeared to employ more controlled feeding patterns, whereas mothers experiencing depression seemingly employed lower levels of controlled feeding.

Conclusions:

Findings that maternal anxiety and depression affect levels of controlled feeding are of particular interest and broadly supportive of the few existing studies. Given the mixed results linking controlled feeding to child BMI reported in previous research, further work is required to determine the relationships between maternal mood, child feeding practices, and BMI.

Introduction

Childhood overweight and obesity are increasingly prevalent. In Australia, 25% of school-age children are overweight or obese,¹ as are 20% of preschoolers.² Childhood obesity is one of the strongest predictors of adult obesity³ and is associated with multiple morbidities, including type 2 diabetes, increased peer victimization, poor-self esteem, and depression.^4,5 Although genetics plays a role,⁶ many risk factors for childhood overweight and obesity have roots in the home environment⁷ through parenting style and parental (especially maternal) eating styles,^8,9 behaviors, and child feeding practices.¹⁰ Accordingly, the role of maternal feeding practices in the etiology of child BMI has been researched, with particular attention directed to the role of controlling feeding practices.^7,11–13 Controlling feeding practices, such as pressuring to eat and restricting and monitoring access to foods, are believed to impede development of children's self-regulation by promoting reliance on external cues of hunger and satiety and to alter the appeal of target foods^14,15 leading to ingestion in the absence of hunger and excessive weight gain.¹⁰ Conversely, pressure to eat reduces preference and intake of healthy target foods.^16,17 Various studies support this connection.^10,18 The occurrence of various parental feeding practices is also known to vary across cultural and socioeconomic groupings^19–21 and may be biased to some extent by child gender.^22,23

However, the relationship between controlling feeding practices and child BMI has not proved robust. Although maternal restriction has shown positive, negative, and null associations with child BMI, the bulk of studies show a positive relationship.^{12,13,24–26} On the other hand, evidence on the relationship between maternal pressure and later child BMI is somewhat inconsistent.²⁰ Interestingly, parental feeding practices characterized by high monitoring uniformly show no association with child BMI.^26–28

A relatively small number of published studies suggest that maternal mood is associated with controlling feeding styles. For example, among mothers of 5-year-old daughters,²⁸ maternal depression was associated with greater restriction and pressure to eat. Furthermore, others have shown this to be mediated by parenting quality.²⁹ Other researchers have reported similar findings regarding parental symptoms of depression, anxiety, and stress.^30,31

One longitudinal study of global maternal psychological distress from pregnancy to 12 months postpartum found no association between mental health (either in pregnancy, or the first year postpartum) and pressure to eat at 1 year postpartum.³² However, maternal use of restriction at 1 year was predicted by anxiety in pregnancy and the postpartum period. Subsequent research revealed that monitoring, restriction, and pressure to eat at 1 and 2 years postpartum were partially predicted by maternal mood at 6 and 12 months postpartum.³³ In short, research on the relationship between maternal mood and controlling feeding practices, while scarce, suggests that an association exists.^32,33

Given the relative scarcity of evidence, the aim of the current study was two fold. The first was to examine whether controlling maternal feeding practices predict child BMI and the second was to investigate whether maternal depressive and anxious symptoms were predictive of controlling child feeding practices. We chose the most widely employed psychometrically sound measure of maternal feeding practices^30,34 and followed a moderately large sample of Australian families longitudinally.

Methods

Participants

A total of 5079 pregnant women were recruited through the Victorian component of The National Postnatal Depression Program, an Australia-wide prospective study into perinatal mental health between 2002 and 2005.³⁵ Women were surveyed between the 6^th and 26^th week of gestation. This time-point is hereafter referred to as Antenatal Time 1. A total of 1252 mothers were approached to complete follow-up data during 2008–2009, hereafter referred to as Early Childhood Time 2. Of these, 223 women responded at Early Childhood Time 2. From these 223 women, data sufficient for analysis were available for only 203 (and from only 159 for analyses involving BMI; see Results section for reasons for this attrition).

Measures

Psychosocial Risk Factor Questionnaire (PSRFQ).³⁶

A psychosocial risk factor questionnaire³⁶ was administered at Antenatal Time 1 and Early Childhood Time 2. The Antenatal Time 1 version, administered as part of the National Postnatal Depression program, obtained demographics, including age, country of birth, language spoken at home, income, occupation, and education. Psychosocial data included previous psychiatric conditions, current emotional/psychological issues, major life events in the past year, availability of emotional/practical support, level of daily hassles, relationships with mother and partner, and history of abuse. The Early Childhood Time 2 version obtained demographic data and psychosocial data analogous to that collected at Antenatal Time 1 but also included the child-specific items of age, gender, current height and weight, schooling, physical health, temperament, language development, anxiety, depression, behavioral disturbances, and parental child-related help seeking.

Body Mass Index

Height and weight were obtained from maternal report on the Early Childhood Time 2 PSRFQ. BMI was computed as weight [kg]/height[m]² and participants were classified as overweight or obese using cutoff criteria for adults and children, respectively.^37,38

The Edinburgh Postnatal Depression Scale.³⁹

The Edinburgh Postnatal Depression Scale (EPDS), a brief 10-item self-rated questionnaire, was administered at Antenatal Time 1. Items cover possible symptoms of depression in the previous 7 days and respondents rate their agreement with each statement on a four-point Likert scale from 0 to 3. Some example items from the scale are: “I have looked forward with enjoyment to things; I have felt sad or miserable; The thought of harming myself has occurred to me.” Total scores are derived through summation of individual items, hence possible total scores range from 0 to 30. The EPDS was designed to screen for postnatal depression and is widely used in research and clinical practice. It has good reliability (α=0.87), has been validated for antenatal use,⁴⁰ and has good acceptability among both depressed and nondepressed women.⁴¹ In an Australian population, the EPDS showed 85% sensitivity, 71% specificity, and 82% overall accuracy for identifying major and minor depression in the postnatal period.⁴² In the best-quality validation studies based on gold-standard diagnostic criteria, the instrument has an average positive predictive value of around 60%.⁴³ The study reported here used a cutoff score of >12 to signify elevated symptoms of depression.

Depression Anxiety Stress Scales 21-item. The Depression Anxiety Stress Scales 21-item⁴⁴ (DASS-21), a 21-item self-rated measure, was used to measure symptoms of depression, anxiety, and stress at Early Childhood Time 2. The DASS was designed to yield continuous measures of symptom severity, in the previous 7 days, on three constructs (depression, anxiety, and stress) in clinical and nonclinical samples. The depression scale focuses on dysphoria, hopelessness, worthlessness, pessimism, self-deprecation, inertia, and anhedonia. (Example item, “I felt that I wasn't worth much as a person.”) The anxiety scale covers physical arousal, situational anxiety and subjective experience of anxious affect (e.g., trembling, pounding heart, apprehension, panic, and worry about loss of control. (Example item, “I was worried about situations in which I might panic and make a fool of myself.”) The stress scale focuses on psychological tension and chronic nonspecific arousal; e.g., inability to relax, tendency to overreact, irritability, and intolerance of interruption or delay. (Example item, “I was intolerant of anything that kept me from getting on with what I was doing.”) In clinical samples, the three scales have good discriminant validity in distinguishing between major depressive disorder and a range of anxiety disorders (panic disorder, generalized anxiety disorder, social phobia, and obsessive compulsive disorder).^45,46 The robust three-factor structure (depression, anxiety, and stress) has been verified by confirmatory factor analyses in several validation studies in clinical and nonclinical populations.^44–47 The three scales of the DASS-21 have good test–retest reliability and internal consistency (r=0.71, 0.79, .081⁴⁵ and α=0.94, 0.87, and 0.91 for the depression, anxiety, and stress scales, respectively⁴⁶).

The Child Feeding Questionnaire. The Child Feeding Questionnaire²⁷ (CFQ), a 31-item self-report questionnaire, was used to measure parental attitudes, beliefs, and practices of child feeding and obesity proneness at Early Childhood Time 2. Item responses are rated on a five-point Likert scale. Composite scores are computed through summation of individual items, forming scores for seven domains: ‘perceived feeding responsibility,’ ‘perceived parent weight,’ ‘perceived child weight,’ ‘concern about child weight,’ ‘restriction,’ ‘pressure to eat,’ and ‘monitoring.’ Only the three controlling feeding subscales namely, ‘restriction’ (e.g., “I have to be sure that my child does not eat too many sweets.”), ‘pressure to eat’ (e.g., “If my child says ‘I am not hungry,’ I try to get him/her to eat anyway.”) and ‘monitoring’ (e.g., “How much do you keep track of the high-fat food that your child eats?”) plus ‘concern for child weight’ (e.g., “I am worried that my son/daughter will become overweight.”) were used in this research. The CFQ has been widely used in clinical and nonclinical groups and shown to correlate with observed feeding interactions.^48,49 It has good internal consistency with the Cronbach α≥0.7 for each scale.²⁸

Procedure

A total of 5079 women completed an Antenatal Time 1 PSRFQ³⁶ and EPDS³⁹ with midwives during routine antenatal visits to maternity hospitals from 2002 to 2005. All participants gave written informed consent. In all, 1252 women were then selected at random and mailed research packages comprising a reply-paid self-addressed envelope, plain language statement, consent form, Early Childhood Time 2 PSRFQ,³⁶ DASS-21,⁴⁴ and CFQ,²⁷ during 2008–2009. Follow-up letters were sent to approximately one-third of nonresponders. Ethics approval for use of the current data was obtained through the Deakin Human Research Ethics Committee and Austin Health Human Research Ethics Committee.

Statistical Analysis

Data were examined for accuracy of entry and missing values. Mean substitution was used for imputation of missing data where <10% of items were missing at random.⁵⁰ Previous work has shown this approach to be robust.⁵¹ Variables were assessed for normality, linearity, homoscedasticity, and univariate outliers. Outliers, identified using a z-score cutoff of 3.29, were recoded to one unit greater or lesser than the next most extreme score.⁵² Mahalanobis distance was used to assess multivariate outliers. t-tests and χ² tests assessed between-group differences on key demographic and PSRFQ variables that might systematically account for participant drop-out and incomplete responding.

The Pearson r measured relationships between continuous variables. Two series of hierarchical multiple regressions were undertaken to explore: (1) The relative contribution of maternal anxiety and depression variables (Antenatal Time 1 EPDS scores and Early Childhood Time 2 DASS-21 scores) in predicting controlling feeding practices and (2) controlled feeding practices in predicting child BMI. Comprising the same steps, the first three models sought to determine predictors of pressure to eat, restriction, and monitoring feeding practices, while controlling for variables previously demonstrating covariance.^53–59 At Step 1, Antenatal Time 1 annual family income, maternal education, and age, Early Childhood Time 2 number of children and child gender were entered. Antenatal depression was entered at Step 2 and maternal Early Childhood Time 2 BMI, concern for child weight and depression, anxiety, and stress (as measured by the DASS-21) were entered at Step 3. The models examining if controlling feeding practices predicted child BMI comprised the same three steps and added an additional step comprising pressure to eat, restriction and monitoring. Analyses were carried out in IBM SPSS Statistics 20 (IBM Corporation, Armonk, NY).

Results

From the 1252 women approached at Early Childhood Time 2, a total of 223 women responded to the invitation to take part in the present study. Data from 5 women were excluded due to: Responding to children born outside the study period, stillborn child, or having twins. Data from an additional 12 women were excluded due to a high frequency (>10%) of missing data and 3 were excluded as multivariate outliers. An additional 44 women failed to provide data on BMI. Hence, data from 203 women were available for analyses not involving BMI and a subsample of 159 women were included in the primary analyses involving BMI.

Descriptive Statistics

Key psychosocial characteristics at Antenatal Time 1 and between-group differences for Time 2 responders and nonresponders are presented in Table 1. Compared to national statistics, mothers in the current sample were slightly more educated.⁶⁰

Table 1.

Characteristics, at Antenatal Time 1, of Time 2 Responders and Time 2 Nonresponders

	T2 Nonresponders (n=1029)		T2 responders (n=203)^a
Variable	Responses (n)	Percentage or mean (SD)	Responses (n)	Percentage or mean (SD)	p value
Age at T1 (years)	1029	30.65 (4.95)	203	33.96 (4.82)	<0.001
Marital status	924		197		0.02
Married		63.7%		72.1%
Not married		36.3%		27.9%
Higher education	684		144		<0.001
Yes		18%		31.7%
No		82%		68.3%
Annual family income	899		199		NS
≤$60, 000		50.0%		48.3%
>$60, 000		27.2%		40.9%
Country of birth	911		198		0.001
Australia		80.9%		90.9%
Elsewhere		19.1%		9.1%
Number of distressing events	905	1.0 (1.23)	197	0.81 (0.99)	0.01
Prior psychiatric diagnosis	623		142		NS
Yes		24.1%		26.8%
No		75.9%		73.2%
Emotional problems in pregnancy	904		196		NS
Yes		28.5%		30.1%
No		71.5%		69.9%
EPDS antenatal score	937	6.62 (4.93)	201	6.20 (4.84)	NS

Note: Annual family income does not add to 100% because varying percentages of participants did not disclose this information. All data were derived from T1 responses.

Only those Time 2 responders with sufficient data for analysis (203 out of 223 responses) are shown in the table.

NS, not significant; EPDS, Edinburgh Postnatal Depression Scale.

For children from the final sample of Time 2 responders (n=203), 54.7% were female at Early Childhood Time 2. Child age ranged from 2.73 to 7.02 years [M=4.82, standard deviation (SD)=0.80]. Among the 159 cases with BMI data, maternal-reported child BMI ranged from 11 to 25 (M=16.42 , SD=2.49), corresponding to 12.6% overweight and 10.1% obese (International Obesity Task Force criteria³⁸) in agreement with recent national statistics showing between 23% to 25% of Australian preschool- to school-aged children are overweight or obese.^61,62

Those responding at both Antenatal Time 1 and Early Childhood Time 2 were significantly older, more likely to be married, and Australian born, and had higher levels of education and fewer distressing events in the prior 12 months than the women completing Antenatal Time 1 data only. At Early Childhood Time 2, those providing full data were significantly older (37 years on average versus 35.1 years for partial responders, p<0.01) and had significantly lower EPDS scores at Early Childhood Time 2 (mean EPDS score=5.4 versus 7.1 for partial responders, p=0.03). In total, 13.8% of this sample scored above cutoff on the EPDS.

Relationships between Key Variables

For mothers of daughters, restriction was correlated positively with concern for child weight (r=0.27, p<0.05), and pressure to eat correlated positively with both the anxiety and depression scales of the DASS-21 at Early Childhood Time 2 (r=0.24 and 0.23, respectively, p<0.05 in both cases). Daughters' BMI values were not associated with feeding patterns or concern for child weight. For mothers of sons, child BMI was associated positively with concern for child weight (r=0.29, p<0.01) and stress was associated positively with restriction (r=0.33, p<0.01). For all mothers, Antenatal Time 1 EPDS scores were correlated positively with Early Childhood Time 2 depression, anxiety, and stress symptoms measured by the DASS-21 (p<0.01 in all cases). Significant positive intercorrelations existed between Early Childhood Time 2 depression, anxiety, and stress for both mothers of daughters and mothers of sons (r>0.5 and p<0.01 in every case).

Prediction of Controlled Feeding

Hierarchical multiple regressions exploring the relative contribution of maternal anxiety and depressive symptoms in predicting controlling feeding practices are presented in Tables, 2, 3, and 4. None of the individual anxiety or depressive symptom variables, or overall models, significantly contributed to the prediction of pressure to eat (Table 2).

Table 2.

Predictors of Maternal Pressure to Eat Feeding Practices

	Model summary				Coefficients
	R	R ²	ΔR²	p value	B	SEB	β
Step 1	0.29	0.09	0.09	0.09
Annual family income					0.00	0.14	0.00
Highest education					−0.21	0.15	−0.14
Mother's age					0.01	0.02	0.03
Number of children					−0.66	0.28	−0.23^*
Child gender					−0.01	−0.19	−0.01
Step 2	0.32	0.10	0.02	0.18
Antenatal EPDS					−0.04	−0.02	−0.19
Step 3	0.37	0.14	0.04	0.57
Mother BMI					−0.01	0.02	−0.05
Concern for child weight					−0.09	0.15	−0.07
DASS-21 Depression scale					0.01	0.02	0.05
DASS-21 Anxiety scale					0.03	0.03	0.17
DASS-21 Stress scale					−0.00	0.02	−0.03

Note: β is the standardized and B is the nonstandardized regression coefficient. N=159 (only those respondents who provided BMI data).

p<0.05.

SEB, Standard Error of B; EPDS, Edinburgh Postnatal Depression Scale; BMI, body mass index; DASS-21, Depression Anxiety Stress Scales 21-item.

Table 3.

Predictors of Maternal Restriction Feeding Practices

	Model summary				Coefficients
	R	R ²	ΔR²	p value	B	SEB	β
Step 1	0.25	0.06	0.06	0.24
Annual family income					−0.07	0.12	−0.06
Highest education					0.27	0.12	0.22^*
Mother's age					0.01	0.02	0.07
Number of children					−0.25	0.23	−0.10
Child gender					−0.01	0.16	−0.01
Step 2	0.25	0.06	0.00	0.89
Antenatal EPDS					−0.07	0.02	−0.11
Step 3	0.40	0.17	0.10	0.06
Mother BMI					0.01	0.02	0.05
Concern for child weight					0.09	0.12	0.07
DASS-21 Depression scale					−0.03	0.02	−0.21
DASS-21 Anxiety scale					−0.01	0.02	−0.05
DASS-21 Stress scale					0.05	0.02	0.46^**

Note: β is the standardized and B is the non-standardized regression coefficient. n =159 (only those respondents who provided BMI data).

p<0.05.

p<0.001.

SEB, Standard Error of B; EPDS, Edinburgh Postnatal Depression Scale; BMI, body mass index; DASS-21, Depression Anxiety Stress Scales 21-item.

Table 4.

Predictors of Maternal Monitoring Feeding Practices

	Model summary				Coefficients
	R	R ²	ΔR²	p value	B	SEB	β
Step 1	0.23	0.05	0.05	0.36
Annual family income					−0.08	0.13	−0.06
Highest education					0.13	0.13	0.11
Mother's age					0.03	0.02	0.18
Number of children					0.07	0.24	0.03
Child gender					0.18	0.17	0.11
Step 2	0.23	0.05	0.00	0.63
Antenatal EPDS					−0.02	0.02	−0.12
Step 3	0.32	0.10	0.05	0.35
Mother BMI					0.01	0.02	0.06
Concern for child weight					0.19	0.13	0.15
DASS-21 Depression scale					−0.02	0.02	−0.14
DASS-21 Anxiety scale					0.01	0.02	0.05
DASS-21 Stress scale					0.02	0.02	.020

Note: β is the standardized and B is the nonstandardized regression coefficient. N =159 (only those respondents who provided BMI data).

SEB, Standard Error of B; EPDS, Edinburgh Postnatal Depression Scale; BMI, body mass index; DASS-21, Depression Anxiety Stress Scales 21-item.

None of the overall models contributed significantly to the prediction of restriction (Table 3). However, the overall model at Step 3 approached near significance [F (11, 99)=1.66, p=0.06], accounting for 10% of the variance in restriction. Among the variables entered at this step, only stress made a significant, independent contribution to the prediction of restriction [t (109)=3.03, p<0.01], but maternal BMI, concern for child weight, and the depression and anxiety scales of the DASS-21 did not. In the prediction of monitoring (Table 4) again, none of the individual anxiety or depressive symptom variables, nor overall models, explained significant amounts of the variance in monitoring.

In the next regression model (Table 5) evaluating the relative contribution of feeding practices to the prediction of child BMI, Step 3 accounted for 10% of the variance in Early Childhood Time 2 child BMI [F (11, 99)=1.99, p<0.05]. However, concern for child weight was the only variable to contribute significant independent variance [t (106)=2.12, p<0.05]. None of the feeding practices contributed to BMI, individually or in combination.

Table 5.

Predictors of Child BMI

	Model summary				Coefficients
	R	R ²	ΔR²	p value	B	SEB	β
Step1	0.26	0.07	0.07	0.21
Annual family income					−0.62	0.38	−0.17
Highest education					−0.27	0.39	−0.07
Mother's age					0.06	0.05	0.10
Number of children					−0.74	0.76	−0.09
Child gender					0.23	0.50	0.04
Step 2	0.28	0.08	0.02	0.18
Antenatal EPDS					0.01	0.06	0.02
Step 3	0.43	0.18	0.10	0.04^*
Mother BMI					0.04	0.05	0.07
Concern for child weight					0.85	0.40	0.21^*
DASS-21 Depression scale					0.09	0.06	0.24
DASS-21 Anxiety scale					−0.09	0.07	−0.19
DASS-21 Stress scale					0.40	0.05	0.12
Step 4	0.45	0.20	0.23	0.42
Pressure to eat					−0.49	0.35	−0.15
Restriction					−0.17	0.26	−0.06
Monitoring					0.33	0.33	−0.11

Note: β is the standardized and B is the nonstandardized regression coefficient. N=159 (only those respondents who provided BMI data).

p<0.05.

SEB, Standard Error of B; EPDS, Edinburgh Postnatal Depression Scale; BMI, body mass index; DASS-21, Depression Anxiety Stress Scales 21-item.

Further regression analyses were conducted among all 203 women responding at Early Childhood Time 2 (the lack of child BMI from 44 women precluded analysis of this variable in the final sample of 203 Time 2 responders). A correlation matrix presenting intercorrelations between variables, stratified by child gender, is presented in Table 6.

Table 6.

Correlations between Maternal Demographics, Psychopathology, Feeding Practices, and Child BMI Stratified by Child Gender in the Larger Research Sample (Females in Top/Right Half and Males in Bottom/Left Half)

	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.	12.	13.	14.
1. Mother's BMI	—	−0.13	−0.02	−0.21^*	−0.12	0.05	−0.09	0.12	0.04	0.06	−0.14	−0.12	0.06	−0.12
2. Mother's age	−0.13	—	−0.02	0.06	−0.07	0.07	0.08	−0.11	−0.08	−0.09	−0.00	−0.02	0.05	−0.06
3. Income	−0.14	0.15	—	0.11	0.12	−0.15	0.21	−0.16	0.01	0.02	−0.23^**	−0.13	−0.24^*	−0.08
4. Education	0.01	−0.06	0.21^*	—	0.05	−0.30^*	0.08	−0.04	0.00	−0.03	0.02	−0.16	−0.19	−0.16
5. No. children	0.03	0.03	−0.01	−0.03	—	0.07	−0.08	−0.16	0.14	−0.01	0.17	−0.08	−0.06	0.00
6. Child BMI	0.13	−0.02	0.01	−0.11	−0.03	—	−0.07	−0.05	0.19	0.15	0.13	0.10	0.11	0.08
7. Restriction	0.02	−0.07	0.08	0.18	0.02	−0.02	—	0.23^*	0.17	0.25^**	0.05	−0.08	−0.07	−0.04
8. PTE	−0.06	−0.03	−0.03	−0.21^*	−0.15	−0.16	0.00	—	0.10	0.00	−0.02	0.17	0.18	0.12
9. Monitoring	0.07	0.09	−0.07	0.06	−0.07	0.02	0.39^**	−0.06	—	0.24^*	−0.01	−0.12	−0.07	0.04
10. CCW	0.25^**	−0.13	−0.08	0.05	−0.09	0.23^**	0.14	−0.04	0.17	—	−0.02	0.02	−0.06	0.08
11. EPDS	0.15	−0.16	0.01	−0.01	0.00	0.07	0.08	−0.07	−0.02	0.14	—	0.25^*	0.29^**	0.41^**
12. Depression	0.02	—	−0.08	−0.11	−0.05	0.17	0.05	0.15	−0.03	0.20^*	0.41^**	—	0.69^**	0.79^**
13. Anxiety	−0.03	−0.21^*	−0.14	−0.18	−0.14	−0.01	0.08	0.08	−0.04	0.17	0.28^**	0.58^**	—	0.65^**
14. Stress	−0.03	−0.13	−0.07	−0.08	−0.06	0.10	0.32^**	0.01	0.04	1.9^**	0.33^**	0.64^**	0.61^**	—
M (Male)	26.55	32.81				16.14	3.35	2.57	4.03	1.29	6.40	5.32	4.39	11.32
SD (Male)	5.37	4.86				2.42	0.88	0.93	0.86	0.63	4.82	6.15	5.54	7.99
M (Female)	27.20	33.15				16.80	3.22	2.52	4.24	1.31	5.97	6.19	3.99	12.07
SD (Female)	5.84	4.81				2.56	0.77	0.96	0.78	0.51	4.89	7.44	4.93	8.47

Note: Income was defined in Australian dollars as an annual family income ≤$60, 000 versus ≥$60, 000. Education was defined as maternal higher education versus no higher education. Number of children was defined as primaparous versus multiparous.

N =203 except in those correlations involving BMI (N=159).

p<05.

p<0.01.

PTE, pressure to eat. CCW, concern for child weight. Items 12, 13 and 14 refer to subscales of the Depression Anxiety Stress Scales 21-item (DASS-21); EPDS, Edinburgh Postnatal Depression Scale; BMI, body mass index; SD, standard deviation.

For mothers of sons, child BMI was associated positively with concern for child weight and restriction was associated positively with stress. For mothers of daughters, concern for child weight was associated positively with restriction and monitoring. However, none of the exploratory models, or the individual anxiety and depression variables, significantly contributed to the prediction of pressure to eat among the larger sample. Nonetheless, a model including maternal BMI, concern for child weight, and the depression, anxiety, and stress scales of the DASS-21 did account for 9% of the variance in restriction [F (11, 188)=2.29, p<0.05]. Concern for child weight [t (187)=2.53, p<0.05], Early Childhood Time 2 depression [t (187)=−2.28, p<0.05], and Early Childhood Time 2 stress [t (187)=3.24, p<0.01], each significantly contributed unique variance to the prediction.

Similarly, a model incorporating Maternal BMI, concern for child weight, depression, anxiety, and stress (as measured by the DASS-21) accounted for 6% of the variance in monitoring [F (11, 188)=1.96]. While concern for child weight was the only variable to significantly independently contribute to monitoring [t (187)=2.51, p<0.05], both the DASS-21 depression scale [t (187)=−1.93, p=0.06] and stress scale [t (187)=1.90, p=0.06] approached significance.

Discussion

The aim of the current study was two-fold: (1) To determine whether maternal antenatal and/or concurrent anxiety and depressive symptoms predict child feeding practices and (2) whether controlling child feeding practices predict child BMI. Maternal depression and anxiety may influence feeding practices through parenting quality, e.g., impaired responsiveness and sensitivity, elevated hostility and control, reduced pleasurable activities, and more sedentary lifestyle.^63–65 These effects have more recently been shown to be mediated via parenting style.²⁹ In this study, neither pressure to eat, restriction, nor monitoring were associated with child BMI after controlling for maternal psychosocial variables. Hence, the evidence for a relationship between controlling feeding practices and child BMI remains equivocal. This is in keeping with other adequately sampled research employing comprehensive, validated assessments of feeding practices, in which results also appear mixed.^{18,30,53,54,57,66–68}

Concern for child weight was associated with restriction. However, an interesting gender difference emerged in this study—the concern appeared largely in mothers of daughters but not mothers of sons. Interestingly, for mothers of daughters, concern for child weight was not associated with child BMI.

The relationship between concern for child weight and more controlling feeding practices is consistent with both prior research^25,67 and with Costanzo and Woody's⁶⁹ domain-specific parenting model, which posits that parents exercise greatest control over feeding when they either: (1) Have their own eating and weight concerns, (2) consider their child as over or underweight, (3) perceive their child at risk of deviant weight, or (4) are concerned about their child's weight.⁶⁹ Possibly, the elevated concern for the weight of daughters may be accounted for by strong Western values, placed particularly on female weight and shape.^30,70–72 Prior research suggests that controlled feeding practices are more frequently implemented by mothers of daughters than mothers of sons.^15,56,69,73

The role of maternal anxiety and depressive symptoms in controlled feeding practices proved complex. In the smaller sample (n=159), the prediction of restriction through the combination of mothers' BMI, concern for child weight, and concurrent maternal depression, anxiety, and stress (as measured by the DASS-21) reached near significance, suggesting these variables are likely influential in promoting mothers use of restriction (together, these accounted for 10% of the variance in restriction). However, stress was the only variable to independently contribute to the prediction after controlling for maternal demographics, suggesting its relative importance in promoting restriction. In the larger sample of Time 2 responders (n=203), stress remained positively associated with restriction in mothers of sons. Interestingly, for mothers of daughters, previously shown associations between pressure to eat and concurrent anxiety and depression (DASS-21) no longer emerged. Regardless, restriction was partially predicted by the combination of mothers' BMI, concern for child weight, and concurrent maternal depression, anxiety, and stress (as measured by the DASS-21). Together, these variables accounted for 9% of the variance in restriction. Interestingly, where concurrent stress contributed positively to the prediction of restriction, concurrent depression contributed negatively such that mothers experiencing heightened depression used less restriction. Among the larger research sample, monitoring was also partially predicted by the combination of mothers' BMI, concern for child weight, and concurrent depression, anxiety, and stress (DASS-21 scales). Although only concern for child weight contributed unique variance to the prediction of monitoring, both concurrent maternal depression and stress approached significance suggesting a likely role. Paralleling predictors of restriction, where concurrent stress positively contributed to variance in monitoring, concurrent depression contributed negatively. This again indicates that women who experienced heightened stress are likely to use high levels of monitoring, whereas women experiencing heightened depression are likely to employ monitoring less.

Restriction

The finding that restriction in particular was partially predicted by maternal distress is consistent with past research.^31,33 The wider parenting literature suggests that during high stress periods, parents tend to exhibit more controlling and uninvolved parental interactions.⁷⁴ Our results suggest that such interactions transfer into the feeding domain. Indeed, heightened stress may impede ability to interpret child cues of satiety, thereby evoking more controlling feeding practices. Furthermore, controlled feeding may temporarily ameliorate feelings of lack of control during high stress periods.⁵⁸ Yet, contrary to prior research^28,30–33 current outcomes showed heightened depression (DASS-21 depression scale) predicted lower levels of controlling feeding styles. Possibly some manifestations of depression may promote more hostile, forceful feeding, while for others it may lead to withdrawn uninvolved interactions.^30,61 It is worth noting that mean depression, anxiety, and stress scores fell within “normal” range for the DASS-21,⁴⁴ suggesting that on average mothers in this study were not experiencing “clinical” levels of anxiety, stress or depression.

Maternal Mood

Contrary to earlier research,^32,33 maternal antenatal depressive symptoms were not predictive of later use of controlling feeding practices. Given the scarcity of longitudinal research tracking maternal depression and anxiety from antenatal to postpartum periods, further investigations are clearly needed before conclusions regarding the role of antenatal maternal psychopathology in predisposing mothers to more controlling feeding practices are drawn. The null associations between antenatal maternal depression (measured by the EPDS) and feeding practices leaves open to question the direction of observed relationships between concurrent maternal anxiety and depressive symptoms and controlled feeding practices. It is possible that controlled feeding practices are a reaction to concern for, or actual, child weight rather than a product of maternal psychopathology; further research longitudinally controlling for child weight is needed to explore this assumption.

Limitations

Outcomes from the current study must be considered in light of several methodological limitations. Data were derived from a largely heterogeneous middle class sample of mothers, who were significantly older, were more likely to be Australian born, married, and to have attained a higher level of education, and experienced fewer distressing events in the prior 12 months, than women completing data at Antenatal Time 1 only. Given that culture,⁷⁵ income,⁵⁷ education,^58,59,71 and maternal age⁵⁸ have been found to influence parental feeding practices, our results need to be interpreted with this proviso in mind. Similarly, on average women in the current sample experienced low levels of depression and anxiety symptoms, so that our findings cannot be generalized to severely depressed or anxious subgroups with certainty. Moreover, data were self-reported by mothers only, and the role of paternal feeding ignored. Conceivably, mothers may have been reluctant to report perceived deviance in child weight or feeding. This may apply particularly to mothers of overweight children for whom feeding is associated with guilt and shame.^57,67 It should further be noted that maternal anxiety and depression symptoms were measured only twice, at two relatively distant time points. Finally, as mentioned earlier, failure to longitudinally track child weight precluded controlling for the same in our analyses. The possibility that the relationship between concurrent maternal anxiety and depression symptoms and controlling feeding practices is a response to child weight cannot therefore be ruled out.

Conclusions

Despite some methodological caveats, the current study adds to research revealing null associations between controlled feeding practices and child BMI even when psychometrically sound assessments of feeding are conducted among a respectably sized sample. Additionally, our findings highlight the role of concern for child weight in promoting controlled feeding practices. Moreover, the current study contributes to the limited longitudinal research on maternal psychopathology and controlling feeding practices. Overall, some support was found for the role of concurrent anxiety and depression symptoms in the prediction of controlled feeding, particularly that characterized by restriction and monitoring. Future research should aim to address the longitudinal effect of concern for child weight in the promotion of controlled feeding and subsequent child BMI. In particular, both the role of paternal influences⁷⁶ and the effect of child gender may be worthy of closer attention.

Footnotes

Acknowledgments

We thank the families who took part in this study. The work was supported by funds from the Austin Medical Research Foundation.

Author Disclosure Statement

All of the authors of this article declare that no competing financial interests exist in relation to this work.

References

Booth

, Dobbins

, Okely

et al. Trends in the prevalence of overweight and obesity among young Australians. Obesity, 2007; 15:1089–1094.

Wake

, Nicholson

, Hardy

et al. Preschooler obesity and parenting styles of mothers and fathers: Australian national population study. Pediatrics, 2007; 120:e1520–e1527.

Power

, Lake

, Cole

. Measurement and long-term health risks of child and adolescent fatness. Int J Obes Relat Metab Disord, 1997; 21:507–526.

Eisenberg

, Neumark-Sztainer

, Story

. Associations of weight based teasing and emotional well-being among adolescents. Arch Pediatr Adolesc Med, 2003; 157:733–738.

Hayden-Wade

, Stein

, Ghaderi

et al. Prevalence, characteristics, and correlates of teasing experiences among overweight vs. non-overweight peers. Obes Res, 2005; 13:1381–1392.

Carnell

, Kim

, Pryor

. Fat brains, greedy genes, and parent power: A biobehavioral risk model of child and adult obesity. Int Rev Psychiatry, 2012; 24:189–199.

Ventura

, Birch

. Does parenting affect children's eating and weight status? Int J Behav Nutr Phys Act, 2008; 5:15.

Cutting

, Fisher

, Grimm-Thomas

et al. Like mother, like daughter: Familial patterns of overweight are mediated by mothers' dietary disinhibition. Am J Clin Nutr, 1999; 69:608–613.

Drucker

, Hammer

, Argas

et al.

Can mothers influence their child's eating behavior?

J Dev Behav Pediatr, 1999; 20:88–92.

10.

Birch

, Fisher

. Mothers' child-feeding practices influence daughters' eating and weight. Am J Clin Nutr, 2000; 71:1054–1061.

11.

Faith

, Storey

, Kral

et al. The feeding demand questionanire: Assessments of parental demands and congnitions concerning parent-child feeding relations. J Am Diet Assoc, 2008; 108:624–630.

12.

Francis

, Birch

. Maternal influences on daughters' restrained eating behavior. Health Psychol, 2005; 24:548–554.

13.

Francis

, Birch

. Maternal weight status modulates the effects of restriction on daughters' eating and weight. Int J Obes Relat Metab Disord, 2005; 29:942–949.

14.

Satter

. Internal regulation and the evolution of normal growth as the basis for prevention of obesity in children. J Am Diet Assoc, 1996; 96:860–864.

15.

Jansen

, Mulkens

, Edmond

et al. From the Garden of Eden to the land of plenty: Restriction of fruit and sweets leads to increased fruit and sweet consumption in children. Appetite, 2008; 51:570–575.

16.

Galloway

, Fiorito

, Francis

et al. ‘Finish your soup’: Counterproductive effects of pressuring children to eat on intake and effect. Appetite, 2006; 46:318–323.

17.

Galloway

, Fiorito

, Lee

et al.

Parental pressure, dietary patterns, and weight status among girls who are ‘picky eaters'

J Am Diet Assoc, 2005; 105:541–548.

18.

Lee

, Mitchell

, Smiciklas-Wright

et al. Diet quality, nutrient intake, weight status and feeding environments of girls meeting or exceeding recommendations for total dietary fat of the American Academy of Pediatrics. Pediatrics, 2001; 107:e95–e102.

19.

Lumeng

, Ozbeki

, Appulgliesse

et al. Observed assertive and intrusive maternal feeding behaviors increase child adiposity. Am J Clin Nutr, 2010; 95:640–647.

20.

Cardel

, Willig

, Dulin-Keita

et al. Parental feeding practices and socioeconomic status are associated with child adiposity in a multi-ethnic sample of children. Appetite, 2012; 58:347–353.

21.

Murashima

, Hoerr

, Hughes

et al. Feeding behaviors of low-income mothers: Directive control relates to lower BMI in children, and a non-directive control relates to a healthier diet in preschoolers. Am J Clin Nutr, 2012; 95:1031–1037.

22.

Agras

, Hammer

, Huffman

et al. Improving healthy eating in families with a toddler at risk for overweight: A cluster randomized controlled trial. J Dev Behav Pediatr, 2012; 33:529–534.

23.

Johannsen

, Johannsen

, Specker

. Influence of parent eating behaviors and child feeding practices on children's weight status. Obesity, 2006; 14:431–439.

24.

Campbell

, Andrianopolous

, Hesketh

et al. Parental use of restrictive feeding practices and child BMI z-score. A 3-year prospective cohort study. Appetite, 2010; 55:84–88.

25.

Gregory

, Paxton

, Brozovic

. Pressure to eat and restriction are associated with child eating behaviour and maternal concern about child weight; but not body mass index in 2-to-4 year-old children. Appetite, 2010; 54:550–556.

26.

Montgomery

, Jackson

, Kelly

et al. Parental feeding style, energy intake and weight status in young Scottish children. Br J Nutr, 2006; 96:1149–1153.

27.

Birch

, Fisher

, Grimm-Thomas

et al. Confirmatory factor analysis of the Child Feeding Questionnaire: A measure of parental attitudes, beliefs and practices about child feeding and obesity proneness. Appetite, 2001; 36:201–210.

28.

Francis

, Hofer

, Birch

. Predictors of maternal child-feeding style: Maternal and child characteristics. Appetite, 2001; 37:231–243.

29.

McConley

, Mrug

, Gilliland

et al. Mediators of maternal depression and family structure on child BMI: Parenting quality and risk factors for child overweight. Obesity, 2011; 19:345–352.

30.

Haycraft

, Blissett

. Controlling feeding practices and psychopathology in a non-clinical sample of mothers and fathers. Eat Behav, 2008; 9:484–492.

31.

Mitchell

, Brennan

, Hayes

et al. Maternal psychosocial predictors of controlling parental feeding styles and practices. Appetite, 2009; 53:384–389.

32.

Farrow

, Blissett

. Is maternal psychopathology related to obesigenic feeding practices at 1 year? Obes Res, 2005; 13:1999–2005.

33.

Blissett

, Farrow

. Predictors of maternal control of feeding at 1 and 2 years of age. Int J Obes Relat Metab Disord, 2007; 31:1520–1526.

34.

Faith

, Scanlon

, Birch

et al. Parent-child feeding strategies and their relationships to child eating and weight status. Obes Res, 2004; 12:1711–1722.

35.

beyondblue. Final Report of the National Postnatal Depression Program, 2001- 2005. 2005 Melbourne: Beyondblue www.beyondblue.org.au/about-us/programs/perinatal-program-beyond-babyblues/national-pnd-research-program. 2013 May 14.

36.

Milgrom

, Gemmill

, Bilszta

et al. Antenatal risk factors for postnatal depression: A large prospective study. J Affect Disord, 2008; 108:147–157.

37.

World Health Organisation. Physical Status: The Use and Interpretation of Anthropometry. WHO: Geneva, 1995.

38.

Cole

, Bellizzi

, Flegal

et al. Establishing a standard definition for child overweight and obesity worldwide: International survey. Br Med J, 2000; 320:1240–1246.

39.

Cox

, Holden

, Sagovsky

. Detection of postnatal depression: Development of a 10 item postnatal depression scale. Br J Psychiatry, 1987; 150:782–786.

40.

Cox

, Holden

. Perinatal Mental Health. A Guide to the Edinburgh Postnatal Depression Scale (EPDS) Gaskell: London, 2003.

41.

Gemmill

, Leigh

, Ericksen

et al. A survey of the clinical acceptability of screening for postnatal depression in depressed and non-depressed women. BMC Public Health, 2006; 6:211.

42.

Milgrom

, Ericksen

, Negri

et al. Screening for postnatal depression in routine primary care: Properties of the Edinburgh Postnatal Depression Scale in an Australian sample. Aust NZJ Psychiatry, 2005; 39:843–849.

43.

Milgrom

, Mendelsohn

, Gemmill

. Does postnatal depression screening work? Throwing out the bathwater, keeping the baby. J Affect Disord, 2011; 132:301–310.

44.

Lovibond

, Lovibond

. Manual for the Depression Anxiety Stress Scales. Psychological Foundation: Sydney, 1995.

45.

Brown

, Chorpita

, Korotitisch

et al. Psychometric properties of the Depression Anxiety and Stress Scale (DASS) in clinical samples. Behav Res Ther, 1997; 35:79–89.

46.

Antony

, Bieling

, Cox

et al. Psychometric properties of the 42-item and 21-item versions of the Depression Anxiety Stress Scales in clinical groups and a community sample. Psychol Assess, 1998; 10:176–181.

47.

Page

, Hooke

, Morrison

. Psychometric properties of the Depression Anxiety Stress Scales in depressed clinical samples. Br J Clin Psychol, 2007; 46:283–297.

48.

Farrow

, Blissett

. Does maternal control during feeding moderate early infant weight gain? Pediatrics, 2006; 118:e293–e298.

49.

Blissett

, Harris

, Kirk

. Effect of growth hormone therapy on feeding problems and food intake in children with growth disorders. Acta Paediatr, 2000; 89:644–649.

50.

Downey

, King

. Missing data in Likert ratings: A comparison of replacement methods. J Gen Psychol, 1998; 125:175–191.

51.

Roth

, Switzer III

, Switzer

. Missing data in multiple item scales: A Monte Carlo analysis of missing data techniques. Organizational Research Techniques, 1999; 2:211–232.

52.

Tabachnick

, Fidell

. Using Multivariate Statistics, 5th. Pearson Education: Boston, 2007.

53.

Baughcum

, Powers

, Johnson

et al. Maternal feeding practice and beliefs and their relationships to overweight in early childhood. J Dev Behav Pediatr, 2001; 22:391–408.

54.

Fisher

, Birch

. Restricting access to foods and children's eating. Appetite, 1999; 32:405–419.

55.

Fisher

, Birch

. Restricting access to palatable foods affects children's behavioural response, foods selection, and intake. Am J Clin Nutr, 1999; 69:1264–1272.

56.

Johnson

, Birch

. Parents' and children's adiposity and eating style. Pediatrics, 1994; 94:653–661.

57.

Kroller

, Warschburger

. Associations between maternal feeding style and food intake of children with a high risk for overweight. Appetite, 2008; 51:166–172.

58.

Hurley

, Black

, Papas

et al. Maternal symptoms of stress, depression and anxiety are related to nonresponsive feeding styles in a statewide sample of WIC participants. J Nutr, 2008; 138:799–805.

59.

Saxon

, Carnell

, van Jaarsveld

et al. Maternal education is associated with feeding style. J Am Diet Assoc, 2009; 109:894–898.

60.

Australian Bureau of Statistics. Education and Work. ABS: Canberra, 2009.

61.

Australian Bureau of Statistics. National Health Survey. ABS: Canberra, 2009.

62.

Commonwealth Scientific and Industrial Research Organisation. 2007. Australian National Children's Nutrition and Physical Activity Survey. CSIRO: Canberra, 2008.

63.

Carter

, Garrity-Rokous

, Chazan-Cohen

et al. Maternal depression and comorbidity: Predicting early parenting, attachment security and toddler social-emotional problems and competencies. J Am Acad Child Adolesc Psychiatry, 2001; 40:18–25.

64.

Lovejoy

, Graczyk

, O'Hare

et al. Maternal depression and parenting behavior: A meta-analytic review. Clin Psychol Rev, 2000; 20:561–592.

65.

van der Bruggen

, Stams

GJJ

, Bogels

. The relation between child and parent anxiety and parental control: A meta-analytic review. J Child Psychol Psychiatry, 2008; 49:1257–1296.

66.

Carnell

, Wardle

. Associations between multiple measures of parental feeding and children's adiposity in United Kingdom preschoolers. Obesity, 2007; 15:137–144.

67.

Gray

, Janicke

, Wistedt

et al. Factors associated with parental use of restrictive feeding practice to control their children's food intake. Appetite, 2010; 55:332–337.

68.

Spruijt-Metz

, Linquist

, Birch

et al. Relation between mothers' child-feeding practices and children's adiposity. Am J Clin Nutr, 2002; 75:581–586.

69.

Costanzo

, Woody

. Domain specific parenting styles and their impact on the child's development of particular deviance: The example of obesity proneness. J Soc Clin Psychol, 1985; 3:425–445.

70.

Fisher

, Birch

. Parents' restrictive feeding practices are associated with young girls' negative self-evaluation of eating. J Am Diet Assoc, 2000; 100:1341–1346.

71.

Tiggemann

, Lowes

. Predictors of maternal control over children's eating behaviour. Appetite, 2002; 39:1–7.

72.

Stearns

. Fat History: Body and Beauty in the West. New York University Press: New York, 1997.

73.

Jordan-Young

. Hormones, context, and “brain gender”: A review of evidence from congenital adrenal hyperplasia. Soc Sci Med, 2012; 74:1738–1744.

74.

Goodman

, Gotlib

. Risk for psychopathology in the children of depressed mothers: A developmental model for understanding mechanisms of transmission. Psychol Rev, 1999; 106:458–490.

75.

Hughes

, Power

, Fisher

et al. Revisiting a neglected construct: Parenting styles in a child feeding context. Appetite, 2005; 44:83–92.

76.

Taylor

, Wilson

, Slater

et al. Parenting and child body mass index: Longitudinal investigation of maternal and paternal influence. Aust J Psychol, 2011; 63:198–206.

Maternal Psychological Factors and Controlled Child Feeding Practices in Relation to Child Body Mass Index

Abstract

Abstract

Background:

Methods:

Results:

Conclusions:

Introduction

Methods

Participants

Measures

Psychosocial Risk Factor Questionnaire (PSRFQ). 36

Body Mass Index

The Edinburgh Postnatal Depression Scale. 39

Procedure

Statistical Analysis

Results

Descriptive Statistics

Relationships between Key Variables

Prediction of Controlled Feeding

Discussion

Restriction

Maternal Mood

Limitations

Conclusions

Footnotes

Acknowledgments

Author Disclosure Statement

References

Psychosocial Risk Factor Questionnaire (PSRFQ).³⁶

The Edinburgh Postnatal Depression Scale.³⁹