Association of Infant Temperament With Subsequent Obesity in Young Children of Mothers With Gestational Diabetes Mellitus

Myles S Faith; James B Hittner; Shanta R Hurston; Jie Yin; Louise C Greenspan; Charles P Quesenberry, Jr; Erica P Gunderson

doi:10.1001/jamapediatrics.2018.5199

. 2019 Mar 11;173(5):424–433. doi: 10.1001/jamapediatrics.2018.5199

Association of Infant Temperament With Subsequent Obesity in Young Children of Mothers With Gestational Diabetes Mellitus

Myles S Faith ^1,^✉, James B Hittner ², Shanta R Hurston ³, Jie Yin ³, Louise C Greenspan ⁴, Charles P Quesenberry Jr ³, Erica P Gunderson ^3,^✉, for the SWIFT Offspring Study Investigators

¹Department of Counseling, School, and Educational Psychology, Graduate School of Education, University at Buffalo–State University of New York, Buffalo

²Department of Psychology, College of Charleston, Charleston, South Carolina

³Division of Research, Cardiovascular and Metabolic Conditions Section, Kaiser Permanente Northern California, Oakland

⁴Department of Pediatrics, Kaiser Permanente, San Francisco, California

Group Information: The SWIFT Offspring Study Investigators are listed at the end of this article.

Accepted for Publication: November 29, 2018.

^✉

Corresponding Author: Myles S. Faith, PhD, Graduate School of Education, University at Buffalo—State University of New York, 420 Baldy Hall, Buffalo, NY 14260 (mfaith@buffalo.edu) and Erica P. Gunderson, PhD, MS, MPH, RD, Division of Research, Kaiser Permanente Northern California, 2000 Broadway, Oakland, CA 94612 (erica.gunderson@kp.org)

Published Online: March 11, 2019. doi:10.1001/jamapediatrics.2018.5199

Author Contributions: Dr Gunderson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Gunderson, Faith.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Faith, Hittner, Hurston, Yin, Gunderson.

Critical revision of the manuscript for important intellectual content: Faith, Hittner, Greenspan, Quesenberry, Gunderson.

Statistical analysis: Faith, Yin, Quesenberry, Gunderson.

Obtained funding: Gunderson.

Administrative, technical, or material support: Hurston, Greenspan.

Supervision: Gunderson.

Conflict of Interest Disclosures: Dr Quesenberry reports having received research funding, unrelated to the current study, from pharmaceutical companies Takeda, Merck & Company Inc, Sanofi-Aventis, Lilly, Genentech, Valeant, and Pfizer. Dr Gunderson reports receiving funding unrelated to the current study from Janssen Pharmaceuticals, Inc. Dr Faith reports receiving funding during the conduct of this study from subcontracts to University at Buffalo from National Institutes of Health grants. No other disclosures were reported.

Funding/Support: This study was funded by a Clinical Research Award from the American Diabetes Association (Dr Gunderson), the National Institute of Child Health and Human Development (grant R01 HD050625; Dr Gunderson), the National Institutes of Diabetes, Digestive, and Kidney Diseases (grant R21 DK103171; Dr Gunderson), and the National Institutes of Health National Center for Research Resources (grant UCSF-CTSI UL1 RR024131).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

The SWIFT Offspring Study Investigators: Erica P. Gunderson, PhD, MPH, MS, RD, Kaiser Permanente Northern California; Kathryn G. Dewey, PhD, Department of Nutrition, University of California, Davis; Myles Faith, PhD, Graduate School of Education, University at Buffalo, State University of New York; Charles P. Quesenberry Jr, PhD, Division of Research, Kaiser Permanente Northern California.

Additional Contributions: Prof Kathryn Dewey, PhD, Department of Nutrition, University of California, Davis, is a coinvestigator on the SWIFT Study and the SWIFT Offspring Study. She contributed to the study design and infant feeding assessment methodology and provided critical input on the manuscript. She was not compensated for her contributions. The authors warmly thank all physicians, clinical staff, research staff, and the SWIFT study participants for their commitment and important contributions to the study.

^✉

Corresponding author.

PMCID: PMC6503510 PMID: 30855657

Key Points

Question

Is infant temperament associated with early childhood obesity risk among offspring of mothers with gestational diabetes mellitus?

Findings

In this cohort study, elevated compared with nonelevated soothability in infants was prospectively associated with more than a 2-fold increased adjusted odds of childhood obesity at ages 2 to 5 years after controlling for numerous potential confounders. This was also associated with early introduction of sugary beverages (at ages <6 months).

Meaning

Among a high-risk population of infants exposed to hyperglycemia in utero, elevated soothability foreshadows early childhood obesity, perhaps in part because caregivers may use sugary drinks to assuage infants.

This cohort study evaluates the association of infant temperament with overweight and obesity status at ages 2 to 5 years in children born to mothers who experienced gestational diabetes mellitus.

Abstract

Importance

Infant temperament is associated with excess weight gain or childhood obesity risk in samples of healthy individuals, although the evidence has been inconsistent. To our knowledge, no prior research has examined this topic among children exposed to gestational diabetes mellitus (GDM) in utero.

Objective

To prospectively evaluate infant temperament in association with overweight and obesity status at ages 2 to 5 years among children born to mothers who experienced GDM.

Design, Setting, and Participants

This prospective cohort study took place at Kaiser Permanente Northern California medical centers. We studied singleton infants delivered at 35 weeks’ gestational age or later to mothers who had been diagnosed with GDM. Data were collected from 2009 to 2016, and data analysis occurred from June 2017 to October 2018.

Exposures

The primary exposures in the child’s first year were soothability, distress to limitations, and activity aspects of temperament, as assessed by a validated questionnaire. Modifiable covariates in the child’s first year included breastfeeding intensity and duration monthly ratio scores, along with the timing of the introduction of sugary beverages and complementary foods.

Main Outcomes and Measures

The primary outcome was child overweight and obesity status, assessed at ages 2 to 5 years. Multinomial logistic regression models estimated adjusted odds ratios and 95% CIs for infants whose temperaments were measured at 6 to 9 weeks of age and categorized as elevated (≥75th percentile) or not elevated in the 3 domains. We controlled for nonmodifiable and modifiable covariates across models.

Results

A total of 382 mother-infant pairs participted, including 130 infants (34.0%) who were non-Hispanic white, 126 infants (33.0%) who were Hispanic, 96 infants (25.1%) who were Asian, 26 infants (6.8%) who were non-Hispanic black, and 4 infants (1.1%) who were of other races/ethnicities. In descriptive analyses, elevated infant soothability and activity temperaments were associated with the early introduction of 100% fruit juice and/or sugar-sweetened beverages (at ages <6 months) and shorter breastfeeding duration (from 0 to <3 months), while elevated distress to limitations was associated with early introduction of complementary foods (at ages <4 months). Elevated soothability consistently was associated with a higher odds of later childhood obesity, with adjusted odds ratios across models ranging from 2.22 (95% CI, 1.04-4.73) to 2.54 (95% CI, 1.28-5.03). Greater breastfeeding intensity and duration (12-month combined) score was associated with lower odds of obesity, independent of infant temperament and other covariates.

Conclusions and Relevance

Among this high-risk population of infants, elevated soothability was associated with early childhood obesity risk, perhaps in part because caregivers use sugary drinks to assuage infants. Soothability temperament may be a novel screening target for early obesity prevention interventions involving responsive feeding and emotion regulation.

Introduction

Children born to mothers who have diabetes during pregnancy are at greater risk of developing obesity, glucose intolerance, insulin resistance, and type 2 diabetes.^1,2,3,4,5,6 Both genetics and the intrauterine environment during fetal development affect future health.⁷ Additionally, higher intensity and longer duration of breastfeeding postnatally are associated prospectively with slower ponderal growth and less weight gain during the first year of life among infants of mothers with gestational diabetes mellitus (GDM).⁸ Curiously, to our knowledge, infant temperament has not been studied in this population for future obesity risk. (Temperament is defined as “individual differences in reactivity and self-regulation that are assumed to have a constitutional basis.”⁹)

In healthy populations, there is emerging but mixed evidence for associations between temperament and child weight status.¹⁰ This is true of soothability, distress to limitations (DTL), and activity temperaments. Soothability reflects a child’s tendency toward “reduction of fussing, crying, or distress when soothing techniques are used by the caregiver.”¹¹^(p72) Wells et al¹² originally showed that less soothable infants had higher skinfold thickness at 2 to 3.5 years of age, although Faith and Hittner¹³ found that high soothability at 1 year of age was associated with a 3.72 increased odds of obesity at age 6 years in the Colorado Adoption Project. That high soothability was associated with increased future odds of obesity is consistent with emerging literature linking greater parental feeding-to-soothe strategies with poorer eating regulation and obesity in children.^14,15 Distress to limitations is “fussing, crying or showing distress while (a) in a confining place or position; (b) in caretaking activities; (c) unable to perform a desired action.”¹¹^(p72) Greater infant DTL was associated prospectively with increased weight for length and body fatness at 3 to 18 months of age in 1 study¹⁴ but not in other research.¹² Finally, activity temperament reflects body movement or motion¹⁶ and has been linked to child weight status in some but not all studies.^17,18 Such inconsistent findings may reflect heterogeneity in methodology or cohort characteristics across investigations.

To our knowledge, no research has evaluated the role of temperament and obesity risk among infants exposed to maternal GDM (ie, hyperglycemia) during fetal life. The purpose of this study is to evaluate the association between 3 aspects of infant temperament (soothability, DTL, and activity levels) and future odds of child overweight and obesity at ages 2 to 5 years among individuals exposed in utero to maternal gestational diabetes. We hypothesized that high soothability, high DTL, and low activity would increase future risk of overweight or obesity in early childhood after controlling for a comprehensive panel of covariates.

Methods

Participants

Participating infants were born to mothers enrolled in the Study of Women, Infant Feeding, and Type 2 Diabetes After Gestational Diabetes (SWIFT).^19,20 The SWIFT Study enrolled 1035 women diagnosed with GDM via the Carpenter and Coustan criteria²¹ who delivered singleton live births at or after 35 weeks’ gestation at Kaiser Permanente Northern California (KPNC) hospitals in 2008-2011 and met other inclusion criteria. (Details were previously published.¹⁹)

The KPNC Institutional Review Board approved all study protocols and materials. Mothers provided written consent for assessments and surveys on infant behaviors and health status.²⁰

Measures

Infant Temperament

Soothability, DTL, and activity were assessed by the validated Rothbart Infant Behavior Questionnaire (IBQ-R).¹¹ Each domain score was calculated as the mean of all items completed for that respective subscale, even if there were missing items, which was common. For the 18-item soothability scale, for example, 67 of 321 participants (20.9%) had no missing items), 110 of 321 participants (34.3%) had 1 to 4 missing items, and 144 of 321 participants (44.9%) had 5 missing items or more. For soothability, the instrument does not generate separate scores for specific parenting strategies (eg, food vs rocking vs screen time) regarding their relative capacity to assuage infant distress.

The IBQ-R was administered when infants were 6 to 9 weeks and 6 months old. We used temperament measures at ages 6 to 9 weeks for risk factor analyses. For children with missing temperament measures at that earlier assessment, scores were imputed using the SAS version 9.4 (SAS Institute) procedure PROC MI to generate multiply imputed data sets. Specifically, 20 imputed data sets were generated for each temperament variable, with separate logistic regression models then conducted for each data set. Results were combined using PROC MIANALYZE to generate valid inferences.

For our logistic regression models, infants were scored as less than the 75th percentile (not elevated) or equal to or greater than the 75th percentile (elevated) for each temperament variable. There are no established national cutoffs for high temperament scores on the IBQ-R. Thus, as recommended on the IBQ-R website, we used sample-specific normative cutoffs.

Anthropometric Measurements of Newborns and Children Ages 2 to 5 Years

We obtained weight and length at birth (2009-2011), and weight and height measurements at well-child outpatient visits at ages 2 to 5 years from the KPNC electronic health records (2011-2016). Weight-for-length z scores at birth were calculated for age and sex groups based on the World Health Organization child growth standards.²² Child body mass index (BMI) ² was obtained from the measurements collected between 2 and 5 years of age. The primary outcome measure was child BMI percentile category at follow-up (ie, normal weight, overweight, or obese). Weight status categories were classified based on age-specific and sex-specific BMI percentiles, which defined normal weight as values less than the 85th percentile, overweight as values between the 85th and less than 95th percentiles, and obesity as values equal to or greater than the 95th percentile.²³ For children with more than 1 well-child visit within this age range, we selected the last pediatric visit with anthropometric measurements that was closest to 5 years of age (range, 24 to 64 months). Of 382 children, 81.4% had weight and height measured at age 52 months or older.

Infant Dietary Intake

Dietary assessment methods are described in detail previously²⁰ and included mailed monthly questionnaires and interviews from 3 in-person examinations (at 6-9 weeks, 6 months, and 12 months of age) in which mothers reported the average daily frequency that infants were fed fruit juices, sweetened juices and drinks, sugar added to water or tea, and any cereal in the bottle or any other foods fed to the child in the past 7 days. Mothers completed a dietary history at 6 months, including offspring age at initiation of specific foods and beverages, as well as the types consumed. From these assessments, we created categorical dietary variables for early vs later initiation of complementary foods and beverages (ie, other than breast-milk or infant formula; ≤4 months vs >4 months) and intake of sugar-sweetened beverages (SSB) and/or 100% fruit juice during the first year of life (<6 months vs ≥6 months or none for SSB and/or 100% fruit juice without added sugar, referred to as sugary beverages).

Breastfeeding Intensity and Duration

Methods for assessing the frequency of breastfeeding and the mean amount and frequency of formula feeding per 7 days to quantify monthly breastfeeding intensity, duration scores, and child age at breastfeeding cessation from monthly mailed questionnaires and surveys at in-person visits have been described previously.⁸ From these assessments, breastfeeding duration was recorded as a continuous measure of total months and categorized as none to 3 months or longer than 3 months. We also calculated a quantitative combined breastfeeding intensity and duration ratio score based on the method of Piper et al,²⁴ yielding a fractional score ranging from 0 (exclusive formula feeding) to 1.0 (exclusive breastfeeding) for each month up to 12 months, as previously described.⁸ These ratios were summed to calculate a total score from birth to 12 months (range, 0 to 12).

Covariates

Covariates recorded for mothers included maternal age; prepregnancy BMI; a 3-hour, 100-g oral glucose tolerance test z score sum; an insulin sensitivity index score based on fasting and a 2-hour, 75-g postload glucose test; insulin levels at 6 to 9 weeks’ postdelivery;²⁵ educational attainment; and racial/ethnic group. Covariates for offspring included gestational age at the mother’s GDM diagnosis, weight-for-length z score at birth, birth order, infant diet, a combined breastfeeding intensity and duration score birth to 12 months, intake of sugary beverages (SSB or 100% fruit juices) during the first year of life, and age at last BMI measurement (at ages 2 to 5 years).

Statistical Analysis

Descriptive statistics are presented as means, SDs, and percentages/proportions. We explored via t tests whether infants with elevated vs not elevated scores on each aspect of temperament differed with respect to infant or mother sociodemographic factors, breastfeeding, maternal health, and infant dietary intake measures at baseline.

To test our hypotheses, a series of nested multinomial logistic regression models were examined for each temperament variable. For all models, the outcome was child weight status at follow-up (ie, normal weight, overweight, or obesity). For the unadjusted logistic model, the sole independent variables were the child temperament variables (soothability, DTL, or activity). For logistic model 1, the independent variables were child temperament plus maternal prepregnancy BMI, maternal age, and maternal 3-hour oral glucose tolerance test z score (to measure GDM severity). This model effectively controlled for potential maternal prepregnancy confounders. Logistic model 2 included all covariates in model 1, plus maternal race/ethnicity, infant weight-for-length z score at birth, and birth order. This model effectively controlled for potential infant confounders at birth. Logistic model 3 included all covariates in model 2, plus infant breastfeeding duration and intensity 12-month score, and age at initiation of sugary beverages (100% fruit juice and/or SSB). This model effectively controlled for potential confounders associated with modifiable feeding behaviors. There was no evidence of child age as a significant effect modifier for any of the temperament measures in the adjusted models. Finally, post hoc analyses explored findings when testing temperament measures as continuous variables.

All analyses were performed using SAS version 9.4 (SAS Institute Inc). P values were obtained from 2-sided tests, with significance defined as P<.05. Covariates were not included if they were not associated with weight status independent of other model covariates (P values >.05) or did not alter 1 or more odds ratios (ORs) for temperament by at least 10% after the addition of covariates. Data analysis occurred from June 2017 to October 2018.

Results

The SWIFT Offspring Study enrolled 466 mother-infant dyads in 2009 to 2011 from the parent SWIFT Study. All infants were healthy, live singleton births born at or after 35 weeks’ gestation without serious medical conditions.²⁰ Of 466 enrolled pairs, 464 mother-infant pairs remained after the exclusion of 2 dropout pairs. Another 31 pairs did not assess infant temperament at 6 to 9 weeks or 6 months of age, and 51 children were missing weight and height measurements in KPNC electronic health records at ages 2 to 5 years. After these exclusions, there were 382 child-mother pairs with assessments and child weight status at ages 2 to 5 years (Figure). A comparison of the final analytic sample vs excluded individuals revealed that mothers who were excluded compared with those whose data were analyzed had only a lower proportion of 4-year college or higher education (28 of 82 [34%] vs 185 of 382 [48.4%]; P = .02; eTable 1 in the Supplement).

Descriptive statistics for all covariates, including missing data information, are provided in Table 1. Of the infants, 130 (34.0%) were non-Hispanic white, 126 (33.0%) were Hispanic, 96 (25.1%) were Asian, 26 (6.8%) were non-Hispanic black, and 4 (1.1%) were of other race/ethnicities. The mean score for temperament domains did not significantly differ by the number of missing items in our sample in post hoc analyses (data not shown). There were no significant differences between temperament scores at ages 6 to 9 weeks and 6 months when comparing children with both values. The mean (SD) age for classification of child weight status was 56.5 (7.0) months, and the median was 59.2 (interquartile range, 53.7-60.8) months.

Table 1. Characteristics of Women With Gestational Diabetes Mellitus and Their Infants, per Infant Temperament Percentile Groups^a^,^b.

Characteristic	Patients, No. (%)
	Entire Cohort (N = 382)	Soothability		Distress to Limitations		Activity Level
	Entire Cohort (N = 382)	<75th Percentile (n = 288)	≥75th Percentile (n = 94)	<75th Percentile (n = 286)	≥75th Percentile (n = 96)	<75th Percentile (n = 288)	≥75th Percentile (n = 94)
Maternal Factors
Maternal age, mean (SD), y	33.3 (4.7)	33.4 (4.7)	33.0 (4.7)	33.3 (4.8)	33.4 (4.2)	33.3 (4.7)	33.4 (4.5)
Education, y
≤ High school	88 (23.0)	64 (22.2)	24 (26)	68 (23.8)	20 (21)	59 (20.5)	29 (31)
Some college	109 (28.5)	78 (27.1)	31 (33)	81 (28.3)	28 (29)	82 (28.5)	27 (29)
≥ 4-year college	185 (48.4)	146 (50.7)	39 (41)	137 (47.9)	48 (50)	147 (51.0)	38 (40)
Race/ethnicity
Non-Hispanic white	111 (29.1)	90 (31.3)	21 (22)	84 (29.4)	27 (28)	94 (32.6)	17 (18)^c
Non-Hispanic black	31 (8.1)	24 (8.3)	7 (7)	21 (7.3)	10 (10)	18 (6.3)	13 (14)
Hispanic	130 (34.0)	95 (33.0)	35 (37)	99 (34.6)	31 (32)	92 (31.9)	38 (40)
Asian	100 (26.2)	72 (25.0)	28 (30)	74 (25.9)	26 (27)	78 (27.1)	22 (23)
Other	10 (2.6)	7 (2.4)	3 (3)	8 (2.8)	2 (2)	6 (2.1)	4 (4)
Parity^d
1	134 (35.1)	97 (33.7)	37 (39)	102 (35.7)	32 (33)	104 (36.1)	30 (32)
2	127 (33.2)	95 (33.0)	32 (34)	100 (35.0)	27 (28)	96 (33.3)	31 (33)
≥3	121 (31.7)	96 (33.3)	25 (27)	84 (29.4)	37 (39)	88 (30.6)	33 (35)
WIC recipient	98 (25.7)	65 (22.6)	33 (35)^e	71 (24.8)	27 (28)	65 (22.6)	33 (35)^f
Prepregnancy BMI
<25	103 (27.0)	75 (26.0)	28 (30)	80 (28.0)	23 (24)	83 (28.8)	20 (21)
25-29.9	110 (28.8)	84 (29.2)	26 (28)	77 (26.9)	33 (34)	83 (28.8)	27 (29)
≥30	169 (44.2)	129 (44.8)	40 (42)	129 (45.1)	40 (42)	122 (42.4)	47 (50)
Prenatal 3-h oral glucose tolerance test, z score sum, mean (SD)	0.0 (2.7)	0.0 (2.7)	−0.1 (2.8)	−0.04 (2.8)	0.1 (2.5)	−0.2 (2.5)	0.6 (3.3)^g
Gestational diabetes mellitus treatment
Diet only	258 (67.5)	195 (67.7)	63 (67)	192 (67.1)	66 (69)	188 (65.3)	70 (74)
Oral hypoglycemic agents	107 (28.0)	79 (27.4)	28 (30)	81 (28.3)	26 (27)	86 (29.9)	21 (22)
Insulin	17 (4.5)	14 (4.9)	3 (3)	13 (4.5)	4 (4)	14 (4.9)	3 (3)
ISI_0,120 scores, mean (SD)	1.7 (0.4)	1.7 (0.4)	1.6 (0.4)	1.7 (0.4)	1.7 (0.5)	1.7 (0.5)	1.6 (0.4)^h
Infant Factors
Sex
Male	207 (54)	153 (53.1)	54 (57)	148 (51.7)	59 (61)	146 (50.7)	61 (65)
Female	175 (46)	135 (46.9)	40 (43)	138 (48.3)	37 (39)	142 (49.3)	33 (35)ⁱ
Birth weight, g	3423 (502)	3406 (477)	3477 (570)	3424 (514)	3422 (464)	3424 (499)	3421 (513)
Birth length, cm	50.7 (2.3)	50.6 (2.2)	50.7 (2.4)	50.7 (2.3)	50.7 (2.3)	50.6 (2.3)	50.9 (2.3)
Size at birth, mean (SD)
Weight-for-length z score	−0.4 (1.3)	−0.4 (1.4)	−0.2 (1.3)	−0.3 (1.3)	−0.4 (1.4)	−0.3 (1.4)	−0.5 (1.3)
Weight-for-age z score	0.2 (1.0)	0.2 (1.0)	0.3 (1.2)	0.2 (1.1)	0.2 (0.9)	0.2 (1.0)	0.2 (1.1)
Length-for-age z score	0.6 (1.2)	0.6 (1.2)	0.6 (1.3)	0.6 (1.2)	0.6 (1.2)	0.6 (1.2)	0.7 (1.2)
Length of stay, mean (SD), d	2.3 (1.4)	2.2 (1.3)	2.4 (1.7)	2.2 (1.5)	2.3 (1.3)	2.3 (1.6)	2.1 (1.0)
Type of nursery
Regular	344 (90.1)	266 (92.4)	78 (83)^j	259 (90.6)	85 (89)	259 (89.9)	85 (90)
Intermediate	19 (5.0)	12 (4.2)	7 (7)	13 (4.5)	6 (6)	12 (4.2)	7 (7)
Neonatal intensive care unit	19 (5.0)	10 (3.5)	9 (10)	14 (4.9)	5 (5)	17 (5.9)	2 (2)
Gestational age at gestational diabetes mellitus diagnosis, mean (SD), wk	25.0 (7.6)	25.4 (7.2)	23.7 (8.7)^k	25.0 (7.6)	25.1 (7.7)	25.0 (7.6)	24.9 (7.8)
Gestational age at birth, mean (SD), wk	39.1 (1.1)	39.1 (1.1)	39.1 (1.1)	39.0 (1.1)	39.2 (1.1)	39.1 (1.1)	39.0 (1.1)
Gestational age, mean (SD), wk
35-36	23 (6.0)	18 (6.3)	5 (5)	19 (6.6)	4 (4)	18 (6.3)	5 (5)
37-39	280 (73.3)	211 (73.3)	69 (73)	209 (73.1)	71 (74)	206 (71.5)	74 (79)
≥ 40	79 (20.7)	59 (20.5)	20 (21)	58 (20.3)	21 (22)	64 (22.2)	15 (16)
Infant feeding at age 6 to 9 wk, mean (SD)
Breastfeeding duration, mo	9.2 (7.0)	9.2 (6.9)	9.3 (7.3)	9.3 (7.3)	9.0 (6.1)	9.6 (7.0)	8.1 (7.0)
Breastfeeding intensity and duration combined score^l	6.4 (4.6)	6.5 (4.6)	6.3 (4.5)	6.4 (4.6)	6.6 (4.5)	6.7 (4.6)	5.5 (4.6)^m
Timing of Initiation of Other Foods
Initiation of solid foods, No./total No.
<6 mo	277/377 (73.5)	215/286 (75.2)	62/91 (68.1)	200/282 (70.9)	77/95 (81)	205/285 (71.9)	72/92 (78)
≥6 mo	100/377 (26.5)	71/286 (24.8)	29/91 (31.9)	82/282 (29.1)	18/95 (19)	80/285 (28.1)	20/92 (22)
Initiation of 100% fruit juice/sugar-sweetened beverages
<6 mo	46/365 (12.6)	28/277 (10.1)	18/88 (20)ⁿ	33/272 (12.1)	13/93 (14)	28/275 (10.2)	18/90 (20)^o
≥6 mo or none	319/365 (87.4)	249/277 (89.9)	70/88 (80)	239/272 (87.9)	80/93 (86)	247/275 (89.8)	72/90 (80)
Initiation of complementary foods and beverages
<4 mo	68/378 (18.0)	51/286 (17.8)	17/92 (18)	42/283 (14.8)	26/95 (27)^p	46/286 (14.7)	22/92 (24)
≥4 mo	310/378 (82.0)	235/286 (82.2)	75/92 (82)	241/283 (85.2)	69/95 (73)	240/286 (83.9)	70/92 (76)
Child age at last BMI measurement, mo	56.5 (7.0)	56.4 (7.0)	56.4 (6.8)	56.5 (6.8)	56.5 (7.6)	56.2 (7.2)	57.3 (6.4)

Open in a new tab

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); ISI_0,120, insulin sensitivity index (from the 2-hour, 75-g oral glucose tolerance test at 6-9 weeks postpartum); WIC, Special Supplemental Nutrition Program for Women, Infants, and Children.

^{^a}

The χ² or Fisher exact test was used for categorical variables; the t test was used for continuous variables.

^{^b}

Missing values: Infant feeding intention score, 13 women; insulin sensitivity index scores, 4 women; initiation of solid foods, 5 infants; initiation of 100% fruit juice/sugar sweetened beverages, 17 infants; initiation of complementary foods, 4 infants.

^{^c}

P = .009 for race/ethnicity by activity category.

^{^d}

The birth order for the child is equal to maternal parity.

^{^e}

P = .02 for WIC participation by soothability category.

^{^f}

P = .02 for WIC participation by activity category.

^{^g}

P = .01 for maternal 2-hour oral glucose tolerance test result by infant activity category.

^{^h}

P = .03 for gestational age at gestational diabetes mellitus diagnosis by soothability category.

^ⁱ

P = .01 for ISI_0,120 result by activity category.

^{^j}

P = .03 for infant sex by activity category.

^{^k}

P = .049 for type of nursery by soothability category.

^{^l}

Sum from birth to 12 months.

^{^m}

P = .02 for breastfeeding intensity and duration 12-month combined score by activity category.

^ⁿ

P = .01 for age at initiation of 100% fruit juice/sugar-sweetened beverages category by soothability category.

^{^o}

P = .01 for age at initiation of 100% fruit juice/sugar-sweetened beverages category by activity category.

^{^p}

P = .006 for age at initiation of 100% fruit juice/sugar-sweetened beverages category by distress-to-limitations category.

Soothability Temperament

Descriptive Statistics and Correlates

Infants high in soothability, compared with those low in soothability, were more likely to have initiated sugary beverages before age 6 months (18 of 88 [20%] vs 28 of 277 [9.7%]; P = .01; Table 1). The percentages of infants initiating sugary beverages at 6 months or older or not at all was smaller for those with high soothability, compared with those with low soothability (70 of 88 [80%] vs 249 of 277 [89.9%]; P = .01). Comparable results were found when categorizing groups as those initiated at ages younger than 6 months of age (18 of 88 [20%] vs 28 of 277 [10.1%]), 6 months or older (40 of 88 [45%] vs 150 of 277 [54.2%]), or never (30 of 88 [34%] vs 99 of 277 [35.7%]; P = .04; eFigure 1 in the Supplement). Additionally, infants with high soothability compared with infants with low soothability had a significantly earlier gestational age at maternal GDM diagnosis (mean [SD] gestational age, 23.7 [8.7] weeks vs 25.4 [7.2] weeks; P = .03) and greater percentages of neonatal intensive care unit admission (9 of 94 [10%] vs 10 of 288 [3.5%]; P = .03) and participation in Special Supplemental Nutrition Program for Women, Infants, and Children (a program of the US government that provides supplemental food, health care referrals, and nutrition education to low-income pregnant women, mothers, and children from birth to age 5 years) (33 of 94 [35%] vs 65 of 288 [22.6%]; P = .02).

Logistic Regression Analyses

High soothability was consistently associated with increased future odds of obesity across the 3 nested logistic regression models (Table 2). Compared with infants who were low in soothability, those in the high-soothability group (≥75th percentile) were more likely to be obese at 2 to 5 years of age in the unadjusted model (odds ratio [OR], 2.54 [95% CI, 1.28-5.03]), model 1 (OR, 2.49 [95% CI, 1.21-5.12]), model 2 (OR, 2.44 [95% CI, 1.17-5.09]), and model 3 (OR, 2.22 [95% CI, 1.04-4.73]). The associations for odds of overweight were somewhat attenuated and nonsignificant. Results of exploratory analyses treating soothability on a continuum revealed statistically significant linear associations for the obesity group only in the unadjusted model (OR, 1.53 [1.05-2.24]; P = .03) and model 1 (OR, 1.54 [1.03-2.29]; P = .03; eTable 2 in the Supplement). There was no evidence for significant quadratic or cubic associations (eTables 3 and 4 in the Supplement). Regarding modifiable feeding covariates, breastfeeding was associated prospectively with reduced obesity risk in model 3 (adjusted OR, 0.91 [95% CI, 0.83-0.98]) and model 4 (fully adjusted OR, 0.90 [95% CI, 0.82-0.98]).

Table 2. Multinomial Logistic Regression Odds Ratios for the Odds of Overweight and Obesity at Ages 2 to 5 Years in Children of Mothers With Gestational Diabetes Mellitus, per Soothability Percentile Groups^a.

Multinomial Logistic Regression Model per Soothability Percentile Groups	Odds Ratios (95% CI) for Child BMI Percentile Groups (N = 382)^b		Overall P Value
	85th-<95th Percentile	≥95th Percentile	Overall P Value
Unadjusted
≥75th	1.45 (0.71-2.93)	2.54 (1.28-5.03)	.008
<75th	1 [Reference]	1 [Reference]	.008
Model 1^c
≥75th	1.43 (0.67-3.05)	2.49 (1.21-5.12)	.01
<75th	1 [Reference]	1 [Reference]	.01
Model 2^d
≥75th	1.48 (0.69-3.19)	2.44 (1.17-5.09)	.02
<75th	1 [Reference]	1 [Reference]	.02
Model 3^e
≥75th	1.44 (0.66-3.13)	2.22 (1.04-4.73)	.04
<75th	1 [Reference]	1 [Reference]	.04

Open in a new tab

Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).

^{^a}

The main independent variable is soothability score at 6 to 9 weeks of age (<75th percentile vs ≥75th percentile).

^{^b}

The normal weight referent is a BMI less than the 85th percentile.

^{^c}

Model 1 includes maternal clinical measures: the sum of prenatal 3-hour 100-g oral glucose tolerance test z scores, prepregnancy BMI, and age in years.

^{^d}

Model 2 combines model 1 with race/ethnicity, birth order (parity), and infant weight-for-length z score at birth.

^{^e}

Model 3 combines model 2 with the sum of breastfeeding intensity and duration ratios (0-1 for each month) from birth through age 12 months; age at initiation of 100% fruit juice/sugar-sweetened beverage intake during the first year (<6 months vs ≥6 months or none; 17 values missing).

Distress to Limitations Temperament

Descriptive Statistics and Correlates

Infants high in DTL, compared with infants low in DTL, more frequently initiated complementary foods prior to 6 months of age (26 of 95 [27%] vs 42 of 283 [14.8%]; P = .006; Table 1). Comparable results were found when categorizing groups as those initiated at ages younger than 4 months (26 of 95 [27%] vs 42 of 283 [14.8%]), 4 to 6 months old (56 of 95 [59%] vs 180 of 283 [63.6%]), and older than 6 months (13 of 95 [14%]) vs 61 of 283 [21.6%]; P = .01; eFigure 2 in the Supplement). Finally, infants high in DTL, compared with those low in DTL, woke up significantly more often at night (median, 2.0 vs 3.0 times; P < .001).

Logistic Regression Analyses

Distress to limitations was not associated with future odds of obesity across models (Table 3). In post hoc analyses, a higher DTL on a continuum was significantly associated with reduced odds of future obesity in model 1 (OR, 0.69 [95% CI, 0.50-0.96]; P = .03; eTable 5 in the Supplement). There was no evidence for significant quadratic or cubic associations (eTables 6 and 7 in the Supplement).

Table 3. Multinomial Logistic Regression Odds Ratios for Overweight and Obesity at Ages 2 to 5 Years in Children of Mothers With Gestational Diabetes Mellitus, per Distress-to-Limitations Percentile Groups^a.

Multinomial Logistic Regression Model Per Distress-to-Limitations Percentile Group	Odds Ratios (95% CI) for BMI Percentile Groups (N = 382)^b		Overall P Value
	85th-<95th Percentile	≥95th Percentile	Overall P Value
Unadjusted
<75th	1.57 (0.70-3.50)	1.09 (0.53-2.26)	.81
≥75th	1 [Reference]	1 [Reference]	.81
Model 1^c
<75th	1.64 (0.71-3.80)	1.10 (0.52-2.34)	.80
≥75th	1 [Reference]	1 [Reference]	.80
Model 2^d
<75th	1.66 (0.71-3.89)	1.07 (0.50-1.07)	0.86
≥75th	1 [Reference]	1 [Reference]	0.86
Model 3^e
<75th	1.70 (0.72-4.04)	1.10 (0.50-2.45)	.81
≥75th	1 [Reference]	1 [Reference]	.81

Open in a new tab

Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).

^{^a}

The main independent variable is the distress-to-limitations score at 6 to 9 weeks of age (<75th percentile vs ≥75th percentile).

^{^b}

The normal weight referent is a BMI less than the 85th percentile.

^{^c}

Model 1 includes maternal characteristics: the sum of prenatal 3-hour 100-g oral glucose tolerance test z score, prepregnancy BMI, and age in years.

^{^d}

Model 2 combines model 1 with maternal race/ethnicity, birth order (parity), and infant weight-for-length z score at birth.

^{^e}

Model 3 combines model 2 with the sum of breastfeeding intensity and duration ratios (0 to 1 for each month) through 12 months of age, plus the age at initiation of juice/sugar-sweetened beverage intake during the first year (<6 vs ≥6 months or none; 17 values missing).

Activity Temperament

Descriptive Statistics and Correlates

Infants who had high activity levels compared with those with low activity levels were significantly more likely to have initiated sugary beverages younger than 6 months (18 of 90 [20%] vs 28 of 275 [10.2%]; P = .01; Table 1). Nonsignificant findings resulted from categorization of groups as younger than 6 months of age (18 of 90 [20%] vs 28 of 275 [10.2%], 6 months or older (43 of 90 [48%] vs 147 of 275 [53.5%]), or never (29 of 90 [32%] vs 100 of 275 [36.4%]; P = .05; eFigure 3 in the Supplement). Infants with high activity, compared with infants with low activity, also had mothers with significantly higher prenatal oral glucose tolerance test results (mean [SD] z score, 0.6 [3.3] vs −0.2 [2.5]; P = .01) and lower postpartum insulin sensitivity index scores (mean [SD], 1.6 [0.4] vs 1.7 [0.5]; P = .01), were less likely to be exclusively breastfed (19 of 94 [20%] vs 86 of 288 [29.9%]) and more likely to be exclusive formula fed (24 of 94 [26%]) vs 34 of 288 [11.8%]; P = .007), and were less likely to be breastfed for more than 3 months (64 of 94 [68%] vs 227 of 288 [78.8%]; P = .03). Finally, infants with high activity were significantly more likely than those with low activity to be male (61 of 94 [65%] vs 146 of 288 [50.7%]; P = .02), have a maternal non-Hispanic black background (13 of 94 [14%] vs 18 of 288 [6.3%]) or maternal Hispanic background (38 of 94 [40%] vs 92 of 288 [31.9%]; P = .01), and be from families who participated in the Special Supplemental Nutrition Program for Women, Infants, and Children program (33 of 94 [35%] vs 65 of 288 [22.6%]; P = .02).

Logistic Regression Analyses

Activity was not associated with future odds of obesity across models (Table 4). In post hoc analyses, higher activity was significantly associated with increased odds of future overweight in model 1 (OR, 1.34 [95% CI, 1.01-1.78]; P = .04) and model 2 (OR, 1.38 [95% CI, 1.03-1.83]; P = .03; eTable 8 in the Supplement); none of the models showed significant associations with future odds of obesity. There was no evidence for significant quadratic or cubic associations (eTables 9 and 10 in the Supplement).

Table 4. Multinomial Logistic Regression Odds Ratios for the Odds of Overweight and Obesity Status at Ages 2 to 5 Years in Children of Mothers With Gestational Diabetes Mellitus by Activity Level Groups at Ages 6 to 9 Weeks and 6 Months^a.

Multinomial Logistic Regression Model per Activity Level Percentile Group	Odds Ratio (95% CI) for BMI Percentile Group (N = 382)^b		Overall P Value
	85th-<95th Percentile	≥95th Percentile	Overall P Value
Unadjusted
≥75th	1.22 (0.60-2.45)	1.40 (0.69-2.82)	.35
<75th	1 [Reference]	1 [Reference]	.35
Model 1^c
≥75th	1.24 (0.59-2.63)	1.26 (0.60-2.66)	.54
<75th	1 [Reference]	1 [Reference]	.54
Model 2^d
≥75th	1.39 (0.64-3.01)	1.34 (0.63-2.87)	.45
<75th	1 [Reference]	1 [Reference]	.45
Model 3^e
≥75th	1.36 (0.62-2.99)	1.13 (0.51-2.50)	.76
<75th	1 [Reference]	1 [Reference]	.76

Open in a new tab

Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).

^{^a}

The main independent variable is the activity level score at 6 to 9 weeks or 6 months of age (<75th percentile vs ≥75th percentile).

^{^b}

The normal weight referent is a BMI less than the 85th percentile.

^{^c}

Model 1 includes maternal characteristics: the sum of prenatal 3-hour 100-g oral glucose tolerance test z score, prepregnancy BMI, and age in years.

^{^d}

Model 2 combines model 1 with maternal race/ethnicity, birth order (parity), and infant weight-for-length z score at birth.

^{^e}

Model 3 combines model 2 with the sum of breastfeeding intensity and duration ratios (0 to 1 for each month) from birth through 12 months of age, plus the age at initiation of 100% fruit juice/sugar-sweetened beverage intake during the first year (<6 vs ≥6 months or none; 17 values missing).

Discussion

Our main finding is that a high soothability temperament in infancy was associated with a 2.2-fold to 2.5-fold increased odds of future obesity at 2 to 5 years of age in children exposed in utero to GDM. This finding was consistent, remaining significant in the context of multiple nonmodifiable and modifiable obesity risk factors. Our findings are consistent with those of Faith and Hittner.¹³ Odds ratios were smaller (approximately 1.5) in post hoc analyses treating soothability on a continuum, although there was no evidence for quadratic or cubic associations of soothability. Thus, there may be a threshold effect of infant soothability and risk of future obesity in childhood.

We also found, to our knowledge for the first time in this population, an association between infant temperament and the initiation of SSB and complementary food consumption. Infants high in soothability or activity temperaments were more likely to initiate intake of fruit juice and/or sugar-sweetened beverages at younger than 6 months. This is important in light of the American Academy of Pediatrics’ recommendation²⁶ that caregivers not introduce fruit juice prior to 12 months of age unless clinically indicated. The association between DTL and early complementary foods is consistent with 1 prior study.²⁷ Because mothers with GDM may use juice and/or complementary foods to soothe, infant emotionality and emotional regulation should be assessed over time and across environments in future research.^28,29 Indeed, infants displaying any emotional distress (compared with none) during home feedings had higher BMIs through age 6 years among Colorado Adoption Project participants.³⁰

Associations between DTL and activity temperaments with risk of future obesity were null (in the main analyses) and/or inconsistent (in the post hoc analyses). This may reflect factors specific to infants of mothers with GDM or other methodological factors and should be reexamined in future research.

We also found that breastfeeding had protective associations with overweight and obesity onset, which is consistent with our previous findings on ponderal growth during the first year of life in this cohort.⁸ Thus, exclusive or predominant breastfeeding may reduce obesity risk for all children, including infants with high soothability, who are more susceptible. We previously showed that high-intensity breastfeeding and longer duration were associated with slower infant ponderal growth and less weight gain compared with formula feeding in the SWIFT Offspring Study cohort.⁸ These findings align with public health recommendations encouraging breastfeeding.

Strengths of this study include the unique prospective cohort of newborns exposed to maternal GDM, the size and racial/ethnic diversity of participants, long-term follow up of outcomes, and comprehensive covariates.

Limitations

Regarding study limitations: first, we did not include observational or physiological measures of temperament. Second, we did not include a direct measure of parent feeding to soothe¹⁵ or parenting self-efficacy.³¹ Third, residual confounding associated with measurement error for beverage and dietary intake covariates may have affected results. Fourth, we did not measure picky eating or emerging eating profiles.³² Fifth, physiological pathways (eg, opioid systems) were not assessed. Sixth, we assessed temperament only at ages 6 to 9 weeks and 6 months, assuming these scores are representative of the full first year of life (ie, the window encompassing breastfeeding and diet assessments). Although temperament is generally stable throughout the first year of life,^33,34 future studies might obtain additional assessments if feasible.

Conclusions

In sum, among infants at risk for developing obesity and type 2 diabetes, those higher in soothability temperament were more likely to become obese when followed up at ages 2 to 5 years. Mechanisms by which high soothability contributes to obesity still need to be identified. Studies of early-life obesity and diabetes prevention might pay greater attention to temperament traits, which parents may recognize from birth more easily than obesity risk. For health professionals, discussions with parents regarding infant soothing strategies, food, and emotional regulation might be considered,³⁵ particularly for infants exposed to GDM in utero. Anticipatory guidance for healthy eating might build from the topic of temperament.^36,37

Supplement.

eTable 1. Characteristics of mother-infant pairs selected for the analytic sample (n=382) and pairs excluded (n=82) among 464 mother-infant pairs enrolled in the SWIFT Offspring Study.

eTable 2. Results of post-hoc multinomial logistic regression models treating soothability as a continuous measure.

eTable 3. Results of post-hoc multinomial logistic regression models treating soothability as a continuous measure and testing for non-linear associations including a quadratic term in the model.

eTable 4. Results of post-hoc multinomial logistic regression models treating soothability as a continuous measure and testing for non-linear associations including quadratic and cubic terms in the model.

eTable 5. Results of post-hoc multinomial logistic regression models treating distress to limitations as a continuous measure.

eTable 6. Results of post-hoc multinomial logistic regression models treating distress to limitations as a continuous measure and testing for non-linear associations including a quadratic term in the model.

eTable 7. Results of post-hoc multinomial logistic regression models treating distress to limitations as a continuous measure and testing for non-linear associations including quadratic and cubic terms in the model.

eTable 8. Results of post-hoc multinomial logistic regression models treating activity as a continuous measure.

eTable 9. Results of post-hoc multinomial logistic regression models treating activity as a continuous measure and testing for non-linear associations including a quadratic term in the model.

eTable 10. Results of post-hoc multinomial logistic regression models treating activity as a continuous measure and testing for non-linear associations including quadratic and cubic terms in the model.

eFigure 1. Percent of children with <6 months, ≥6 months, and no intake of sugary (fruit juice and/or sugar-sweetened) beverages among soothability groups (p-value= 0.035).

eFigure 2. Percent of children with <4 months, 4-6 months, and ≥6 months initiation of complementary foods among distress to limitations groups (p-value= 0.013).

eFigure 3. Percent of children with <6 months, ≥6 months, and no intake of sugary (fruit juice and/or sugar-sweetened) beverages among activity groups (p-value= 0.051).

Click here for additional data file.^{(280.3KB, pdf)}

References

1.Dabelea D, Knowler WC, Pettitt DJ. Effect of diabetes in pregnancy on offspring: follow-up research in the Pima Indians. J Matern Fetal Med. 2000;9(1):83-88. doi: [DOI] [PubMed] [Google Scholar]
2.Franks PW, Looker HC, Kobes S, et al. Gestational glucose tolerance and risk of type 2 diabetes in young Pima Indian offspring. Diabetes. 2006;55(2):460-465. doi: 10.2337/diabetes.55.02.06.db05-0823 [DOI] [PubMed] [Google Scholar]
3.Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal gestational diabetes, birth weight, and adolescent obesity. Pediatrics. 2003;111(3):e221-e226. doi: 10.1542/peds.111.3.e221 [DOI] [PubMed] [Google Scholar]
4.Petitt DJ, Bennett PH, Knowler WC, Baird HR, Aleck KA. Gestational diabetes mellitus and impaired glucose tolerance during pregnancy: long-term effects on obesity and glucose tolerance in the offspring. Diabetes. 1985;34(suppl 2):119-122. doi: 10.2337/diab.34.2.S119 [DOI] [PubMed] [Google Scholar]
5.Silverman BL, Rizzo T, Green OC, et al. Long-term prospective evaluation of offspring of diabetic mothers. Diabetes. 1991;40(suppl 2):121-125. doi: 10.2337/diab.40.2.S121 [DOI] [PubMed] [Google Scholar]
6.Silverman BL, Rizzo TA, Cho NH, Metzger BE; The Northwestern University Diabetes in Pregnancy Center . Long-term effects of the intrauterine environment. Diabetes Care. 1998;21(suppl 2):B142-B149. [PubMed] [Google Scholar]
7.Dabelea D, Pettitt DJ. Intrauterine diabetic environment confers risks for type 2 diabetes mellitus and obesity in the offspring, in addition to genetic susceptibility. J Pediatr Endocrinol Metab. 2001;14(8):1085-1091. doi: 10.1515/jpem-2001-0803 [DOI] [PubMed] [Google Scholar]
8.Gunderson EP, Greenspan LC, Faith MS, Hurston SR, Quesenberry CP Jr; SWIFT Offspring Study Investigators . Breastfeeding and growth during infancy among offspring of mothers with gestational diabetes mellitus: a prospective cohort study. Pediatr Obes. 2018;13(8):492-504. doi: 10.1111/ijpo.12277 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Rothbart MK, Ahadi SA, Evans DE. Temperament and personality: origins and outcomes. J Pers Soc Psychol. 2000;78(1):122-135. doi: 10.1037/0022-3514.78.1.122 [DOI] [PubMed] [Google Scholar]
10.Bergmeier H, Skouteris H, Horwood S, Hooley M, Richardson B. Associations between child temperament, maternal feeding practices and child body mass index during the preschool years: a systematic review of the literature. Obes Rev. 2014;15(1):9-18. doi: 10.1111/obr.12066 [DOI] [PubMed] [Google Scholar]
11.Gartstein MA, Rothbart MK. Studying infant behavior via the Revised Infant Behavior Questionnaire. Infant Behav Dev. 2003;26:64-86. doi: 10.1016/S0163-6383(02)00169-8 [DOI] [Google Scholar]
12.Wells JC, Stanley M, Laidlaw AS, Day JM, Stafford M, Davies PS. Investigation of the relationship between infant temperament and later body composition. Int J Obes Relat Metab Disord. 1997;21(5):400-406. doi: 10.1038/sj.ijo.0800420 [DOI] [PubMed] [Google Scholar]
13.Faith MS, Hittner JB. Infant temperament and eating style predict change in standardized weight status and obesity risk at 6 years of age. Int J Obes (Lond). 2010;34(10):1515-1523. doi: 10.1038/ijo.2010.156 [DOI] [PubMed] [Google Scholar]
14.Stifter CA, Anzman-Frasca S, Birch LL, Voegtline K. Parent use of food to soothe infant/toddler distress and child weight status: an exploratory study. Appetite. 2011;57(3):693-699. doi: 10.1016/j.appet.2011.08.013 [DOI] [PubMed] [Google Scholar]
15.Stifter CA, Moding KJ. Understanding and measuring parent use of food to soothe infant and toddler distress: a longitudinal study from 6 to 18 months of age. Appetite. 2015;95:188-196. doi: 10.1016/j.appet.2015.07.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Slining MM, Adair L, Goldman BD, Borja J, Bentley M. Infant temperament contributes to early infant growth: a prospective cohort of African American infants. Int J Behav Nutr Phys Act. 2009;6:51. doi: 10.1186/1479-5868-6-51 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Li R, O’Connor L, Buckley D, Specker B. Relation of activity levels to body fat in infants 6 to 12 months of age. J Pediatr. 1995;126(3):353-357. doi: 10.1016/S0022-3476(95)70447-7 [DOI] [PubMed] [Google Scholar]
18.Carey WB. Temperament and increased weight gain in infants. J Dev Behav Pediatr. 1985;6(3):128-131. doi: 10.1097/00004703-198506000-00006 [DOI] [PubMed] [Google Scholar]
19.Gunderson EP, Matias SL, Hurston SR, et al. Study of Women, infant feeding, and type 2 diabetes mellitus after GDM pregnancy (SWIFT), a prospective cohort study: methodology and design. BMC Public Health. 2011;11(1):952. doi: 10.1186/1471-2458-11-952 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Gunderson EP, Hurston SR, Dewey KG, et al. The study of women, infant feeding and type 2 diabetes after GDM pregnancy and growth of their offspring (SWIFT Offspring study): prospective design, methodology and baseline characteristics. BMC Pregnancy Childbirth. 2015;15:150. doi: 10.1186/s12884-015-0587-z [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Expert Committee on the Diagnosis and Classification of Diabetes Mellitus Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 2000;23(suppl 1):S4-S19. [PubMed] [Google Scholar]
22.WHO Multicentre Growth Reference Study Group WHO child growth standards based on length/height, weight and age. Acta Paediatr Suppl. 2006;450:76-85. [DOI] [PubMed] [Google Scholar]
23.Ul-Haq Z, Mackay DF, Fenwick E, Pell JP. Meta-analysis of the association between body mass index and health-related quality of life among children and adolescents, assessed using the pediatric quality of life inventory index. J Pediatr. 2013;162(2):280-6.e1. doi: 10.1016/j.jpeds.2012.07.049 [DOI] [PubMed] [Google Scholar]
24.Piper S, Parks PL. Use of an intensity ratio to describe breastfeeding exclusivity in a national sample. J Hum Lact. 2001;17(3):227-232. doi: 10.1177/089033440101700306 [DOI] [PubMed] [Google Scholar]
25.Gutt M, Davis CL, Spitzer SB, et al. Validation of the insulin sensitivity index (ISI(0,120)): comparison with other measures. Diabetes Res Clin Pract. 2000;47(3):177-184. doi: 10.1016/S0168-8227(99)00116-3 [DOI] [PubMed] [Google Scholar]
26.Heyman MB, Abrams SA; Section on Gastroenterology, Hepatology, And Nutrition; Committee on Nutrition . Fruit juice in infants, children, and adolescents: current recommendations. Pediatrics. 2017;139(6):e20170967. doi: 10.1542/peds.2017-0967 [DOI] [PubMed] [Google Scholar]
27.Wasser H, Bentley M, Borja J, et al. Infants perceived as “fussy” are more likely to receive complementary foods before 4 months. Pediatrics. 2011;127(2):229-237. doi: 10.1542/peds.2010-0166 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Kogan N, Carter AS. Mother-infant reengagement following the still-face: the role of maternal emotional availability in infant affect regulation. Infant Behav Dev. 1996;19:359-370. doi: 10.1016/S0163-6383(96)90034-X [DOI] [Google Scholar]
29.Leerkes EM, Wong MS. Infant distress and regulatory behaviors vary as a function of attachment security regardless of emotion context and maternal involvement. Infancy. 2012;17(5):455-478. doi: 10.1111/j.1532-7078.2011.00099.x [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Hittner JB, Johnson C, Tripicchio G, Faith MS. Infant emotional distress, maternal restriction at a home meal, and child BMI gain through age 6 years in the Colorado Adoption Project. Eat Behav. 2016;21:135-141. doi: 10.1016/j.eatbeh.2016.01.008 [DOI] [PubMed] [Google Scholar]
31.Anzman-Frasca S, Stifter CA, Paul IM, Birch LL. Infant temperament and maternal parenting self-efficacy predict child weight outcomes. Infant Behav Dev. 2013;36(4):494-497. doi: 10.1016/j.infbeh.2013.04.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Hittner JB, Faith MS. Typology of emergent eating patterns in early childhood. Eat Behav. 2011;12(4):242-248. doi: 10.1016/j.eatbeh.2011.06.001 [DOI] [PubMed] [Google Scholar]
33.Bornstein MH, Donnellan MB, Scaramella LV, Hahn CS, Auestad N, O’Connor DL. Infant temperament: stability by age, gender, birth order, term status, and SES. Child Dev. 2015;86(3):844-863. doi: 10.1111/cdev.12367 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Carnicero JC, Perez-Lopez J, Salinas M, Martinez-Fuentez MT. A longitudinal study of temperament in infancy: stability and convergence of measures. Eur J Pers. 2000;14(1):21-37. doi: [DOI] [Google Scholar]
35.Saltzman JA, Pineros-Leano M, Liechty JM, Bost KK, Fiese BH; STRONG Kids Team . Eating, feeding, and feeling: emotional responsiveness mediates longitudinal associations between maternal binge eating, feeding practices, and child weight. Int J Behav Nutr Phys Act. 2016;13:89. doi: 10.1186/s12966-016-0415-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Cameron JR, Rice DC. Developing anticipatory guidance programs based on early assessment of infant temperament: two tests of a prevention model. J Pediatr Psychol. 1986;11(2):221-234. doi: 10.1093/jpepsy/11.2.221 [DOI] [PubMed] [Google Scholar]
37.Cameron JR, Rice DC, Sparkman G, Neville HF. Childhood temperament-based anticipatory guidance in an HMO setting: a longitudinal study. J Community Psychol. 2013;41(2):236-248. doi: 10.1002/jcop.21526 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials