This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Overweight is a major health issue, and parent-targeted interventions to promote healthy development in children are needed.
The study aimed to evaluate E-health4Uth Healthy Toddler, an intervention that educates parents of children aged 18 to 24 months regarding health-related behaviors, as compared with usual care. The effect of this intervention on the following primary outcomes was evaluated when the children were 36 months of age: health-related behaviors (breakfast daily, activity and outside play, sweetened beverage consumption, television (TV) viewing and computer time), body mass index (BMI), and the prevalence of overweight and obesity.
The BeeBOFT (acronym for breastfeeding, breakfast daily, outside playing, few sweet drinks, less TV viewing) study is a cluster randomized controlled trial involving 51 Youth Health Care (YHC) teams. In total, 1094 parents participated in the control group, and 1008 parents participated in the E-health4Uth Healthy Toddler intervention group. The intervention consisted of Web-based personalized advice given to parents who completed an eHealth module and discussion of the advice during a regular well-child visit. In this study the eHealth module was offered to parents before two regular well-child visits at 18 and 24 months of age. During the well-child visits, the parents’ personalized advice was combined with face-to-face counseling provided by the YHC professional. Parents in the control group received usual care, consisting of the regular well-child visits during which general information on child health-related behavior was provided to parents. Parents completed questionnaires regarding family characteristics and health-related behaviors when the child was 1 month (inclusion), 6 months, 14 months, and 36 months (follow-up) of age. The child’s height and weight were measured by trained health care professionals from birth through 36 months of age at fixed time points. Multilevel linear and logistic regression models were used to evaluate the primary outcomes at 36 months of age.
At 36 months, we observed no differences between health-related behaviors of children, BMI or the percentage of children having overweight or obesity in the control and intervention group (
The E-health4Uth Healthy Toddler intervention resulted in small improvements in health-related behaviors among subgroups but had no significant effects with respect to the children’s BMI. We conclude that the E-health4Uth Healthy Toddler intervention may be useful for pediatric health care professionals in terms of providing parents with personalized information regarding their child’s health-related behaviors.
Netherlands Trial Register: NTR1831; http://www.trialregister.nl/trialreg/admin/rctview.asp?TC=1831 (Archived by WebCite at http://www.webcitation.org/6mm5YFOB0)
In 2009, 1.8% of the Dutch boys and 2.2% of the Dutch girls in the age group of 2 to 21 years were classified as being obese, and 12.8% of the boys and 14.8% of the girls were classified as being overweight [
In the Netherlands, Youth Health Care (YHC) is a free program that monitors each child’s health and development and helps families promote healthy behaviors and prevent disease. These benefits are offered to parents and their children in the form of appointments with YHC at set intervals beginning in the child’s first year of life; in total, parents are offered a maximum of 11 well-child visits per child. During a well-child visit, growth and development of the child is assessed and discussed using standardized measures and protocols. Although voluntary, approximately 95% of parents in the Netherlands participate in this program [
The E-health4Uth Healthy Toddler intervention is based on the following theories of behavior change: social-ecological theories and models, including the theory of planned behavior [
The aim of this study was to compare the effects of applying the E-health4Uth Healthy Toddler intervention versus usual care (control) by assessing the following primary outcomes: breakfast daily, activity and outdoor play, sweetened beverages, screen time (ie, television (TV) watching and/or computer use), BMI, and the prevalence of overweight/obesity [
In 2009, 50 preventive YHC organizations in the Netherlands were invited to participate in a 3-armed cluster randomized controlled trial (RCT) entitled the BeeBOFT (acronym for breastfeeding, breakfast daily, outside playing, few sweet drinks, less TV viewing) study (Netherlands Trial Register: NTR1831) [
Within each of these ten organizations, the teams were randomly assigned to one of the following three groups using a computerized random allocation generator: the control group (n=17 teams), the E-health4Uth Healthy Toddler intervention group (n=17 teams), and the BeeBOFT + intervention group (n=17 teams). The E-health4Uth Healthy Toddler intervention group invited parents, at the child age of 18 and 24 months, to complete a Web-based eHealth module providing tailored health education regarding healthy child nutrition and activity behaviors and to discuss this advice during the regular well-child visit with a YHC professional. Therefore, Internet literacy was an implicit eligibility criterion. The YHC teams allocated to the BeeBOFT + intervention group focused on effective child rearing by parents from birth onwards by enlarging parental skills concerning healthy behavioral lifestyle habits. Parents in the control group received usual care, consisting of the regular well-child visits during which general information is provided with regard to health and development of the child. In this study, we focus only on the effects of the E-health4Uth Healthy Toddler intervention compared with usual care.
It is important to note that the YHC professionals and parents were not blinded with respect to the groups. The research proposal was reviewed by the Medical Ethics Committee of the Erasmus University Medical Center. On the basis of their review, the Committee concluded that the Dutch Medical Research Involving Human Subjects Act (in Dutch: Wet medisch-wetenschappelijk onderzoek met mensen) did not apply to this research proposal. The Medical Ethics Committee therefore had no objection to the execution of this study (proposal number MEC-2008-250). Further details regarding the study design and interventions have been published previously [
From January 2009 through September 2010, the 51 participating YHC teams invited the parents of 7985 children to participate in the BeeBOFT study during their first YHC well-child visit, which was conducted at the parents’ home approximately 2-4 weeks after the birth of the child. Parents were requested to provide written informed consent to participate in the 3-year study. In total, 3003 parents agreed to participate in the BeeBOFT study and provided written informed consent (a participation rate of 37.61%; 3003/7985). At inclusion, a questionnaire was completed by the parents; 3 participants did not complete this questionnaire. This questionnaire contained items regarding the pregnancy (eg, duration and complications), childbirth (eg, complications and height and weight at birth), and family demographics (eg, the country of birth of both parents and grandparents, parents’ education level, and number of siblings).
When the child was 6, 14, and 36 months of age, all parents participating in the BeeBOFT study were invited to complete a more extensive questionnaire containing items regarding the child’s health-related behaviors, determinants of these behaviors, and the child’s health-related quality of life (HRQoL). The questionnaires could be completed on paper or online by either the mother or the father. The response rates at the three ages were 77.62% (2331/3003), 77.20% (2318/3003), and 73.46% (2206/3003), respectively.
Here, we present our analysis of the effects comparing children in the E-health4Uth Healthy Toddler intervention group (n=1008 parents) and the control group (n=1094 parents) on the primary outcomes measured when the children were 36 months of age. The results obtained comparing children in the BeeBOFT + intervention group (n=991 parents) and the control group will be reported elsewhere. An overview of the YHC teams and the study participants is presented in
The E-health4Uth Healthy Toddler intervention has been described in detail by Raat et al [
Parents allocated to the intervention group received an invitation to visit the E-health4Uth Healthy Toddler intervention website [
Flow of participants.
As a first step, parents completed the assessment questionnaire; the questions in this questionnaire were based on previous research [
After reading the tailored advice, parents could make an implementation-intention plan in which they could specify actions (eg, what, when, and where to improve child health-related behavior). The tailored advice and implementation plan was sent by email to the parents. During the subsequent well-child visit, the advice was discussed with the YHC professional. The YHC professional, with permission of the parents, also had access to the advice parents received. During these well-child visits at approximately 18 and 24 months of age, the YHC professionals prescribed intervention conditions based on motivational interviewing techniques to help parents change their child’s health-related behaviors [
The intervention software (TailorBuiler) was developed by OverNite Software Europe (OSE, Geleen, the Netherlands).
In the control group, the parents received usual care, which included regular YHC well-child visits. The YHC professionals in the control group provided care in accordance with the YHC Overweight Prevention Protocol [
Health-related behaviors were assessed using the BeeBOFT study questionnaires. All questions were adapted from Dutch questionnaires that were used in previous studies [
Daily breakfast was assessed by asking how many days of the week the child ate breakfast. Parents were instructed to report how many days per week and how much time per day their child spent being active and playing outdoors. Activity and outdoor play were added up to calculate an average hours of activity per day. Parents were instructed to indicate the number of glasses of sweetened beverages their child drank per weekday and weekend day. Examples of sweetened beverages were provided. Daily consumption in glasses per day was calculated. Finally, TV viewing and computer use were assessed by asking parents to report the average number of hours their child spent watching TV and/or using the computer per weekday and weekend day. Screen time in hours per day was calculated by adding up TV viewing time and computer time. At each YHC well-child visit, the child’s height and weight were measured in accordance with standardized protocols [
The following measures were included to describe the study population, evaluate potential confounding factors [
The eHealth module contained questions regarding usability of the module, including the ease of use and whether information regarding health-related behaviors was conveyed in a pleasant manner.
To compare the characteristics between the intervention group and the control group, we used either the Student’s t-test (for continuous variables) or the chi-square test (for categorical variables) [
Consistent with the data analyses described by Raat et al [
Here, we present the results of three regressions models. The first model did not include a correction for cluster (YHC team); the second model included corrections for cluster; and the third model included corrections for cluster and covariates [
The following variables were evaluated as potential confounders: (1) pregnancy duration and birth weight; (2) maternal BMI; (3) maternal HRQoL and well-being (ie, self-rated health); and (4) maternal age, education level, and ethnic background [
Residuals followed a skewed distribution, and health-behavior assessments were log transformed for both the baseline and follow-up variables. Daily breakfast was dichotomized into daily breakfast yes/no because of non-normal distributed residuals after log transformation. Log transformation was performed using the natural logarithm. A constant of 0.01 was added because of zero values; activity n=1, sweet beverages n=192, and screen time n=25. The effect of research group can be interpreted by exponentiation of the coefficient as an approximate percent change in the outcome; percent change in the intervention group=100 x (exp[B]−1).
Subsequently, we evaluated moderation of the intervention effect by child’s gender and ethnic background, as well as the mother’s education level and weight status [
In addition to the data analyses described in the study protocol [
To gain insight into the characteristics of the study participants who were lost to follow-up, we used descriptive statistics to compare the age, country of birth, living situation, and education levels of the mothers who remained in the study through the follow-up measurement (n=1543) with the mothers who were lost to follow-up (n=559). In addition, to get a deeper understanding of the characteristics of the mothers who completed the eHealth module, we also evaluated the abovementioned characteristics between the mothers who completed the eHealth module (n=651) and the mothers who did not complete the eHealth module (n=357) when the child was 18 months of age.
Descriptive statistics were analyzed using the Statistical Package for the Social Sciences (SPSS) version 21.0 (IBM Corp). Generalized linear mixed models and other linear mixed models were performed using Statistical Analysis Software (SAS) version 9.4 (SAS Institute Inc).
Characteristics of the study population at study inclusion (n=2102 parents).
Characteristics | Total study |
Control group |
Intervention group |
|||
Male | 1048 (50.60) | 567 (52.55) | 481 (48.49) | .07 | ||
Age at inclusion in months (missing n=15), mean (SDb) | 0.47 (0.80) | 0.55 (0.95) | .03 | |||
Age at the follow-up assessment in months (missing n=810), mean (SD) | 36.87 (2.47) | 36.81 (2.27) | .67 | |||
Birth weight in kilogramsc (missing n=14), mean (SD) | 3458.41 (525.73) | 3453.60 (528.41) | 3463.61 (523.03) | .66 | ||
Dutch | 1713 (81.65) | 862 (78.94) | 851 (84.59) | .001 | ||
Both parents | 2029 (97.92) | 1053 (97.68) | 976 (98.19) | .44 | ||
Age in years (missing n=28), mean (SD) | 30.88 (4.33) | 30.99 (4.41) | 30.75 (4.24) | .21 | ||
Pregnancy duration in days (missing n=66), mean (SD) | 277.38 (10.72) | 277.01 (10.96) | 277.77 (10.44) | .11 | ||
Yes | 1584 (76.60) | 795 (73.75) | 789 (79.70) | .002 | ||
BMIe in kg/m2 (missing n=152), mean (SD) | 25.09 (4.21) | 25.26 (4.45) | 24.91 (3.88) | .06 | ||
The Netherlands | 1931 (92.17) | 984 (90.36) | 947 (94.14) | .001 | ||
.26 | ||||||
Low | 272 (13.12) | 148 (13.77) | 124 (12.43) | .15 | ||
Mid | 725 (34.97) | 359 (33.40) | 366 (36.67) | .80 | ||
High | 1076 (51.91) | 568 (52.84) | 508 (50.90) | .07 | ||
Employed | 1683 (82.70) | 867 (82.3) | 816 (83.18) | .60 | ||
Very good or excellent | 790 (56.88) | 395 (57.92) | 395 (55.87) | .45 |
aThe
bSD: standard deviation.
cBirth weight was collected by the Youth Health Care professional; if missing, this value was obtained from the parent’s inclusion questionnaire.
dEthnic background of the child was based on the grandparents’ country of birth as described by Statistics Netherlands. If one or both grandparent were born outside the Netherlands, the parents were categorized as non-Dutch. If one or both of the parents were categorized as non-Dutch, the child was also categorized as being of non-Dutch origin.
eBMI: body mass index.
fSelf-rated health of the parent when the child was 36 months of age.
Descriptive summary of the primary outcomes measured at 14 months and at follow-up.
Primary outcomes | 14 monthsa | 36 monthsb | |||||
Control group | Intervention group | Control group | Intervention group | ||||
Daily breakfast 7 days/week, % | 98.0 | 98.0 | .55 | 96.7 | 98.3 | .03 | |
Activity, hours/day, mean (SDd) | 1.91 (1.30) | 1.88 (1.24) | .72 | 2.56 (1.40) | 2.68 (1.13) | .19 | |
Sweetened beverages, glasses/day, mean (SD) | 1.34 (1.16) | 1.39 (1.24) | .39 | 2.31 (1.51) | 2.10 (1.28) | .003 | |
Screen timee, hours/day, mean (SD) | 1.22 (0.92) | 1.05 (0.74) | <.001 | ||||
BMIf, mean (SD) | 16.75 (1.25) | 16.83 (1.24) | .18 | 15.66 (1.29) | 15.78 (1.23) | .12 | |
BMI-SDSg, mean (SD) | -0.25 (0.96) | -0.17 (0.94) | .10 | -0.17 (1.02) | -0.06 (1.01) | .048 | |
Overweight or obesityh, % | 3.99 | 4.77 | .51 |
aNumber of missing values range 448 to 541
bNumber of missing values range 559 to 915
c
dSD: standard deviation.
eNot assessed before 36 months.
fBMI: body mass index.
gBMI-SDS: body mass index-standard deviation score.
hPercentage of overweight and obesity defined by the international age-and gender specific cutoff values; cannot be defined before the age of 24 months.
At 36 months of age, significantly more children in the intervention group ate breakfast daily (
At 36 months of age, the BMI of children in the control group and the intervention group was 15.66 (SD 1.29) and 15.78 (SD 1.23), respectively (
The results of the regression analyses are summarized in
With regard to BMI the third model (ie, the model corrected for potential covariates) revealed a beta value for BMI at follow-up of .10 (95% CI −0.15 to 0.36) for the children in the intervention group as compared with the children in the control group; for BMI-SDS, beta was .12 (95% CI −0.091 to 0.33). With respect to overweight/obesity, the odds ratio at 36 months for the children in the intervention group was 0.79 (95% CI 0.44-1.43) compared with the children in the control group.
In addition, we analyzed BMI and BMI-SDS longitudinally. The interaction term between the study group and age was not significant for either BMI or BMI-SDS (
Results of the three models evaluating primary outcomes among the children at 36 months of age.
Primary outcomes at 36 months | Model 1a,b | Model 2c,b | Model 3d,b | |||
Intervention groupe | Intervention groupe | Intervention groupe | ||||
Daily breakfast (yes), ORe (95% CI)f | 1.55 (0.74 to 3.25) | .25 | 1.55 (0.72 to 3.34) | .25 | 1.31 (0.56 to 3.10) | .52 |
Activity and outdoor playg, hours/day, beta (95% CI) | .04 (−0.02 to 0.10) | .18 | .04 (−0.05 to 0.13) | .38 | .05 (−0.04 to 0.15) | .29 |
Sweetened beveragesg, glasses/day, beta (95% CI) | −.14 (−0.31 to 0.03) | .11 | −.14 (−0.31 to 0.03) | .11 | −.16 (−0.34 to 0.03 | .10 |
Screen timeg, hours/day, beta (95% CI) | −.14 (−0.23 to −0.04) | .005 | −.09 (−0.27 to 0.08) | .30 | −.07 (−0.25 to 0.12) | .47 |
BMIh, beta (95% CI) | .11 (−0.03 to 0.25) | .13 | .11 (−0.15 to 0.36) | .40 | .10 (−0.15 to 0.36) | .43 |
BMI-SDSi, beta (95% CI) | .11 (0.00 to 0.22) | .06 | .12 (−0.09 to 0.33) | .26 | .12 (−0.09 to 0.33) | .28 |
Overweight or obesityj, OR (95% CI) | 0.83 (0.48 to 1.46) | .52 | 0.83 (0.46 to 1.49) | .52 | 0.79 (0.44 to 1.43) | .43 |
aModel 1: corrected for the previous assessment of the outcome (where available).
bModels evaluating BMI, BMI-SDS and % overweight or obesity are corrected for birth weight of the child.
cModel 2: corrected for cluster Youth Health Care (YHC) team and the previous assessment of the outcome (where available).
dModel 3: corrected for cluster (YHC team), the previous assessment of the outcome (where available), the child’s ethnic background, and the child’s precise age at follow-up.
eOR: odds ratio.
fThe estimated coefficients and their 95% confidence interval (95% CI) are given for the children in the intervention group relative to the children in the control group.
gThe previous assessment of the outcome (where available) and the outcome at follow-up were log transformed
hBMI: body mass index.
iBMI-SDS: body mass index-standard deviation score.
jPercent overweight or obese is based on the definition reported by Cole et al [
After observing a significant interaction term between potential moderators and the study groups, we performed stratified analyses. Our analysis revealed that the boys in the intervention group were more likely to eat breakfast daily compared with the boys in the control group at follow-up (OR 10.20; 95% CI 1.75-88.60). Non-Dutch children in the intervention group were 25.86% more active at follow-up compared with the non-Dutch children in the control group (95% CI 0.80-56.83,
The eHealth module was completed primarily by the mother when the children were 18 and 24 months of age (626/651, or 96.2%, and 610/638 missing n=2, or 95.9%, respectively). The parents also reported that they found the eHealth module easy to use (470/651 missing n=81, or 82.5%, and 469/638 missing n=67, or 82.1%, when the children were 18 and 24 months of age, respectively). At 18 months, 60.1% (342/651 missing n=82) and 61.6% (350/651 missing n=83) of parents appreciated receiving information regarding physical activity and nutrition, respectively, via the eHealth module.
On an average, the mothers who participated through to the follow-up time point were older than the mothers who were lost to follow-up (31.11, SD 4.18 vs 30.24, SD 4.65 years, respectively;
In total, 651 out of 1008 (64.58%) parents completed the eHealth module when their child was 18 months of age, compared with 357 out of 1008 (35.42%) parents who did not complete the eHealth module at this time point. On an average, the mothers who completed the eHealth module at this time point were older than the mothers who did not complete the module (31.06, SD 4.06 vs 30.19, SD 4.50 years, respectively;
In this study, we evaluated the effects of the E-health4Uth Healthy Toddler intervention on the child’s health-related behaviors and BMI [
This study adds to the overall knowledge base regarding educating parents in order to optimize the healthy behaviors of young children [
Our results show that at baseline and at 36 months of age, the children in the intervention group had a higher BMI as compared with the children in the control group; we currently have no explanation for this finding. In fact, compared with the age- and gender-matched reference population measured in 1980, the children in the entire study sample had relatively healthier BMI-SDS values. It is important to note that the Dutch reference population values from 1980 were measured just before the overweight epidemic [
The importance of factors in early life that determine the development of overweight among children reflects the need to further develop and optimize interventions designed specifically for parents of young children [
Other elements of the E-health4Uth Healthy Toddler intervention can be optimized and/or revised. For example, to increase its effectiveness, the intervention can be revised to give specific advice to parents of children beyond 24 months and/or to at an earlier age (ie, before 18 months) [
The E-health4Uth Healthy Toddler intervention offers the opportunity to provide parents with important messages that are strengthened by personal counseling with the YHC professional [
Future studies should be designed to test potential beneficial effects of combining an eHealth module with face-to-face counseling among various subgroups such as children of less-educated parents, children of non-Dutch ethnic background, and gender subgroups. Therefore, the E-health4Uth Healthy Toddler intervention can be easily adapted for use in these subgroups [
The strength of our approach was our collaboration with the Dutch YHC organizations, which enabled us to perform a large-scale cluster RCT using Web-based eHealth combined with face-to-face counseling by community-based pediatric health care professionals engaged in daily practice. This approach also provided the opportunity to obtain a large dataset of objectively measured height and weight outcomes for the children whose parents participated in the study. In addition, the response rate among the parents was relatively high (approximately 75%), enabling us to conduct a relatively thorough evaluation of the intervention’s effects on health-related behaviors and BMI.
A limitation of the study was the transformation of outcome variables, namely activity, sweetened beverages, and screen time. We performed sensitivity analyses using the dichotomized version of these primary outcome variables (data not shown) and similar results, that is, no significant differences between intervention and control group, were observed. Another possible limitation of this study was the use of self-report questionnaires, in which parents may have underestimated or overestimated their child’s behavior. However, given that the same assessment materials were used in both the intervention and control groups, this effect—if present—would not likely have affected our results. Moreover, parents with a higher education were more likely to participate in the follow-up measure and in the intervention. Even though the sample may not perfectly represent the general population in the Netherlands, the sample size was sufficiently large for us to evaluate the potential moderating effects among higher and lower educated parents. Regardless, future research and implementation of these types of interventions should emphasize on hard-to-reach lower educated parents. One option for reaching these parents is to provide the eHealth module in the waiting room when the parents arrive for their well-child visit; this approach would also create an opportunity for parents to ask questions to the nurse and/or physician directly after completing the eHealth module.
This large cluster RCT evaluated the E-health4Uth Healthy Toddler intervention, which combines an eHealth module with face-to-face interaction between parents and YHC professionals. Our analysis revealed limited evidence with regard to health behavior and overweight prevention in young children. However, some indications for effects among subgroups of parents and children, such as less-educated parents, were observed. The E-health4Uth Healthy Toddler intervention is relatively easy to implement in community medicine and preventive pediatric practice and can serve as an important addition to current medical guidance and health-promoting practices. This low-intensity intervention can be added to regular care and may save health care professionals valuable time that can be used to focus on health-related behaviors that are determined to be most relevant by the eHealth module. The combination of personalized advice and face-to-face counseling likely increases the effectiveness of this type of intervention [
E-health4Uth Healthy Toddler intervention example advise sweetened beverages.
E-health4Uth Healthy Toddler intervention example advise activity.
Summary of the items for assessing the children’s health-related behaviors.
Overview of interaction terms.
Splined average of the BMI values in the intervention and control group.
CONSORT E-HEALTH checklist (V1.6.1).
body mass index
body mass index-standard deviation score
health-related quality of life
odds ratio
randomized controlled trial
Statistical Analysis Software
standard deviation
Statistical Package for the Social Sciences
television
youth health care
This study was funded by a grant from ZonMW, the Netherlands Organization for Health Research and Development (grant number 50-50110-96-491) and a grant from NWO, the Netherlands Organization for Scientific Research. These funding sources had no role in the design or conduct of the study; in the collection, analysis, or interpretation of the data; or in the preparation, review, approval, or submission of the manuscript.
None declared.