Early childhood obesity has reached alarming proportions. Recent estimates suggest that 8.1% of children from birth to 2 years in the United States have excess weight for their length.1 Unfortunately, special populations are at an even greater risk to experience this health disparity. Children with Down syndrome (DS) are more likely to be obese than their peers who are developing typically.2 Mounting evidence suggests that preventative efforts during pregnancy and early life represent a period in development most sensitive to alter risk for obesity.3 Comprehensive interventions that include physical activity (PA) as an outcome have proven effective in addressing excess weight gain.4 Although interventions targeting obesity shed light on the modifiable factors associated with adiposity, very little is known regarding typical PA patterns experienced by infants. In particular, there is a paucity of evidence on PA patterns in special populations, including infants with DS. A greater understanding of PA patterns during infancy will enable interventionists to more accurately prescribe preventative methods to curb obesity trajectories early on in development.
Recent research demonstrates that elevated levels of childhood body mass index are associated with an increased risk for morbidities as an adult.5 An unhealthy weight status as a child has been shown to contribute to metabolic diseases and obesity-related morbidity during adulthood.4 Most concerning, however, is that a clear health disparity emerges when considering special populations. Children with DS are at an increased risk for hyperinsulinemia and dyslipidemia, both of which are highly influenced by obesity.6 Owing to the host of negative health outcomes associated with obesity, modifiable factors contributing to this epidemic are at the forefront of research priorities. Physical inactivity has been cited as a major contributor to the obesogenic environment of the current population.7
The International Classification of Functioning model emphasizes the goal of participation.8 Within this model, interventions should attempt to measure participation especially related to PA. Regular PA participation in children with DS is difficult because of a variety of constraints, and as a result this population experiences fewer opportunities to engage in health-enhancing PA.9 In a study describing the PA patterns in children with DS and their siblings who are developing typically (aged 3-10 years), children with DS participated in significantly fewer minutes of vigorous PA.10 Furthermore, cross-sectional research supports that children with DS (aged 8-15 years) are less active as they increase in age.11 These findings underscore the need to create interventions early in development when trajectories may be most sensitive to change, particularly in special populations where behavioral patterns are more stable and therefore increasingly challenging to alter.
One of the most cited public health messages is that prevention is critical, meaning interventions should begin in the first 2 years of life.12 Recent research suggests that PA should be included in the comprehensive preventative treatment of obesity during early infancy.4 In fact, PA in the prevention of childhood obesity was recently recommended as a research priority for infants.12 Although evidence-based PA recommendations do not yet exist, the Let's Move Initiative,13 SHAPE America,14 and the Canadian Physical Activity Guidelines15 recommend that infants spend multiple supervised periods of the day on their tummy to promote the achievement of early movement milestones. Despite these initiatives, there are very few evidence-based physical active interventions to implement during infancy.
In one study, the effect of a preventative prone positioning intervention in infants with and without DS was evaluated.16 Findings suggests that to promote the achievement of early motor milestones and PA, between 60 and 75 minutes per day should be spent on tummy time in prone position.16 The most promising finding from this study was, for infants receiving the intervention, Ponderal index decreased following the intervention.16 In a separate study examining the effect of a low- and high-intensity treadmill training study in infants with DS, infants enrolled in the high-intensity training exhibited increased levels of activity at the trunk and ankle compared with infants in the low-intensity group.17 Infants with DS tend to reach the onset of independent walking approximately 12 to 16 months later than their peers who are developing typically.18,19 These studies represent the limited, but important, interventions that include PA as an outcome. Promoting PA before delays emerge may reduce health disparity throughout development.
The effect of PA on the prevention of overweight and obesity has not been adequately studied. A review of PA in young children reported that higher levels of activity during infancy were associated with lower skinfold thickness, greater motor development, and increased cognitive development.20 A recent study using advanced inertial movement sensors with both triaxial accelerometry and gyroscope functions found that infants who produced more leg movements walked later than less active infants.21 However, the authors noted the difficulty in explaining this delay with a small sample.21 The scope of PA research in infants is very narrow, provides only low- to moderate-quality evidence, includes small sample sizes, and is reliant on parental questionnaires or direct observation methods, which are limited in accurately estimating daily activity levels for this age group.20,22 Objective measurement of PA in infants via accelerometry has been proposed as the ideal measurement method,20,22 but work is needed to develop consistent procedures that are appropriate for infants.
To contribute to the growing body of research suggesting prevention during infancy has a key role in curbing early trajectories of obesity, a baseline of PA during this period must first be identified. Next, examining PA during infancy in special populations may lend support for interventions promoting activity patterns early in development. Therefore, the aim of this study was to evaluate the PA patterns in infants developing typically and to compare the results with a group of infants with DS. A greater understanding of the potentially modifiable factors contributing to the obesity epidemic will provide important therapy prescription information for infants with DS.
The Health Sciences and Behavioral Sciences Institutional Review Board at the University of Michigan approved all study methods and procedures. Recruitment was initiated through local DS chapters as well as parent and infant child classes in South East Michigan. Caregiver and infant dyads were invited to participate in the study if the infant was between 1 and 12 months of age and not independently walking.
Demographic information including race, gender, and age was collected through a parent survey. The Bayley Scale of Infant Development III (BSID)23 was administered to each infant participant to measure motor development. The BSID is a standardized assessment developed to measure gross and fine motor skills. The BSID was selected, as it represents a motor assessment that is composed of skills and equipment common to children ranging from 1 to 42 months of age. Quotients were used to describe the motor results in both groups. Quotients were calculated by summing basal- and ceiling-level scores on each represented subscale. The BSID was administered and scored by 3 researchers with extensive experience in motor assessments in young children with and without developmental disabilities. Ponderal index (kg/m3) was calculated from measurements of infant weight and length. Ponderal index is less correlated with length than body mass index, making it a preferable measure of leanness among infants.24 Length-for-age, height-for-age, and weight-for-length percentiles were also calculated on the basis of World Health Organization growth standards.25
PA was measured using accelerometers (Actigraph GT3X+, Pensacola, Florida). Although the form of PA may be different in infants from older children, the movement captured by accelerometry is consistent with definitions of increasing energy expenditure above resting levels.20 Among infants, these movements would include moving arms and legs via kicking, grasping, reaching, pulling, pushing, and tummy time.15 Parents were asked to place 2 monitors, one on the infant's wrist and one on the ankle, for 7 days during all daytime hours including sleep. Accelerometers were secured to an elastic band equipped with a plastic buckle and covered by a lightweight washable cloth cover, similar to receiving blanket material.
Although actigraphy has been accepted as the most valid and reliable estimate of PA in young children,26 there is little published data on the use of accelerometers in infants. One factor contributing to the paucity of literature is the challenge in addressing the volume of external motion, such as adult handling or mechanical movement of the infant.27 To document and remove external motion, caregivers completed a monitoring log. They were instructed to classify their infant's PA behavior in 30-minute intervals during daytime hours for each of the 7 days. Caregivers were provided 4 categories to select from, which included (A) nonwear time when a monitor was removed, (B) adult handling or mechanical movement because of holding, stroller, or an infant sling, (C) quiet play such as seated activity, or (D) active play such as tummy time or kicking activity. Data from the log assisted researchers in evaluating the total amount of PA attributable to movements produced solely by the infant and not because of external motion. Accelerometers and parental logs were returned by priority mail after the 7-day period.
Accelerometer Data Reduction
Actilife 6 software (Pensacola, Florida) was used to download the accelerometer data from wrist and ankle monitors. All data were analyzed with a 15-second epoch length, consistent with recommendations for young children.26 Because classification of PA levels using cut points (ie, sedentary, light, moderate, and vigorous PA) is not available for infants,22,26 all PA data are expressed in average counts per minute (CPM) on the basis of the vector magnitude of the triaxial accelerometer. In general, higher average CPM represents a more active infant.
A standardized protocol was developed to reduce the actigraphy data. The purpose of the protocol was to remove periods of nonwear and external motion from the data set. Nonwear time including periods of bathing, swimming, or misplacement of the monitors was removed on the basis of the parental log. Periods with external motion including adult handling or mechanical movement were removed on the basis of the parental log. Periods with at least 2 consecutive minutes of zeroes were removed and were not counted as wear time. To accurately identify the start of each day, 3 consecutive minutes of activity were identified within the first 30-minute interval recorded by the parent. Wear time for the day was calculated beginning after this first 3-minute period and included all daily hours not removed during processing. This cleaning procedure was used consistently across the study with a high interrater reliability (intraclass correlation coefficient 2,1 ≥ 0.95).
Statistical analyses were conducted using SPSS 22.0 (IBM Corp, Armonk, New York) software with an a priori α of 0.05. Frequencies, proportions, means, and standard deviations were used to describe the characteristics of the sample. Differences between infants with DS and those developing typically were examined through independent samples t tests for age and Pearson's χ2 test for gender, age groupings, and race.
Analyses of covariance (ANCOVA) were used to examine differences between groups for measures of accelerometry, body composition, and motor development. Covariates in each analysis included age and gender. Actigraph wear time was also included as a covariate in analysis involving accelerometry variables. Effect sizes including partial eta-squared (η2) adjusted for estimated marginal means and Cohen's d from unadjusted group means were also calculated. To examine differences between accelerometry at the ankle and wrist, paired samples t tests were conducted separately for each group. Age was a significant covariate in most analyses. ANCOVA analyses examined differences in accelerometry CPM across age groups by months, controlling for gender and wear time. Post hoc pairwise comparisons with Bonferroni corrections examined differences between cross-sectional age groups. Analyses across age were conducted separately for infants with DS and those developing typically.
Linear regression analyses were used to identify possible relationships between accelerometer CPM and gender, age, Ponderal index, BSID motor composite, Actigraph wear time, and presence of DS. Semi-partial (sr) and squared (sr2) correlations were calculated to examine the magnitude of variance explained by individual factors. Regression analyses were conducted separately for accelerometry at the ankle and wrist.
Table 1 includes the characteristics of the study sample. No significant differences were identified between or within groups across gender, age, and race (P > .05). Infants had usable data from accelerometers on an average of 9.27 (standard deviation = 2.62) hours per day and on an average of 6.71 (standard deviation = 0.66) days of the week. There were no differences in usable wear time between groups or placements (P > .05).
Table 2 includes the differences in PA between groups. No significant differences were identified in CPM at the ankle (P = .296, d = 0.075) or wrist (P = .171, d = 0.151) between infants with DS and those developing typically. Infants with DS produced significantly more PA counts at the wrist (469.75 ± 41.00 CPM) than the ankle (332.98 ± 26.23 CPM) location (t(10) = 2.84, P = .018, d = 1.07). Significantly greater wrist CPM was also observed for infants who were developing typically (t(21) = 10.13, P < .001, d = 1.21). Age was a significant covariate for both ankle (P < .001; η2 = 0.585) and wrist (P < .001; η2 = 0.400) CPM.
Table 3 includes the differences in body composition and motor development between groups. No group differences were observed for most measures of body composition (P > .05). However, length-for-age percentile was significantly lower among infants with DS (P = .011; d = 0.983). Age was a significant covariate for Ponderal index (P = .031; η2 = 0.151), but no other measures of body composition (P > .05). Indices of motor development were significantly different between groups (P < .001). Infants with DS exhibited poorer gross motor skills (d = 1.55), fine motor skills (d = 1.45), and gross motor composite (d = 1.59) compared with infants developing typically. Age was a significant covariate for the motor composite (P = .035; η2 = 0.144) and gross motor skills (P = .043; η2 = 0.134), but not for fine motor skills (P = .100; η2 = 0.091).
To examine differences in accelerometer counts in cross-sectional age groups, additional ANCOVA analyses were conducted across groups by months of age. Figure 1 presents the estimated marginal means and standard errors of counts across age groups, controlling for gender and wear time. Among infants developing typically, there was a univariate effect of age on counts at both the ankle (F (5,15) = 5.25, P = .006, η2 = 0.636) and wrist (F (5,14) = 6.78, P = .002, η2 = 0.708). No significant univariate effects of age were observed among infants with DS (ankle: F (5,3) = 6.71, P = .074, η2 = 0.918; wrist: F (5,3) = 0.18, P = .953, η2 = 0.230); however, the effect at the ankle location trended toward statistical significance (P= .074). Pairwise comparisons identified significant differences between the 9- to 10-month and 11- to 12-month groups compared with the 1-to 2-month groups among infants developing typically (P < .05). The sample size of the group with DS limits statistical power and the ability to detect significant post hoc differences (βankle = 0.52; βwrist = 0.06).
Table 4 shows the associations of accelerometry with demographics, body composition, and motor development in infants. For both ankle and wrist CPM, there were significant associations with age and motor development. Age was the factor with the strongest association with PA counts at the ankle (β = 0.851, P < .001) and wrist (β = 0.724, P < .001). Squared semi-partial correlations (sr2) suggest that age independently accounts for 50% (sr2 = 0.50) and 35% (sr2 = 0.35) of the variance in CPM at the ankle and wrist, respectively. Motor development also had significant associations with CPM at the wrist (β = 0.556, P = .027), and trended toward significance at the ankle (β = 0.391, P = .066). Independent contributions to total variance were substantial at both the wrist (sr2 = 0.09) and ankle (sr2 = 0.05), but were considerably smaller than the contribution of age. Once again, the presence of DS was not significantly associated with accelerometry results (P > .05).
This study is one of the first attempts to describe the daily patterns of PA in infants while addressing issues of adult handling or mechanical movements. Understanding the PA patterns of infants is important for addressing early onset of obesity in childhood and will help to inform early interventions. After manually removing periods of nonwear and external motion from the actigraphy data, infants who are developing typically produced an average of 367.59 and 541.03 CPM at the ankle and wrist, respectively. We believe this represents the typical pattern of movement for an infant when averaged across a 7-day period. Within our sample, PA patterns were not significantly associated with adiposity as measured by Ponderal index. However, it is still probable that infant activity may be negatively correlated with adiposity later in childhood.20
Despite the psychological and physical benefits derived from regular PA, children with DS have been reported to participate in less activity than children without DS.28 The results of the present study would suggest that PA patterns are not significantly different between infants with DS and those developing typically during the first year of life. However, average CPMs were lower at both the ankle and wrist for infants with DS. This disparity is likely to expand throughout childhood and adolescents.6,28 An important next research priority would be to examine the PA trajectories in infants and young children with DS, to determine when PA patterns depart from peers who are developing typically. Knowledge gained from this research could assist service providers in the introduction of interventions to address this health disparity during a sensitive time in development.
A further consideration regarding the results from this study is the differences in PA observed between the early and late months. Given the significant changes in motor development and therefore expanded repertoire of motor skills, an increase in self-produced movements would be expected. Therefore, it was not surprising that PA was significantly associated with age and that differences in PA were found between 9 to 10 and 11 to 12 months of age when compared with PA at 1 to 2 months of age. Infants with DS and those developing typically both produced significantly more PA at the wrist compared with the ankle. These results are consistent with the progression of growth and development that suggests a cephalocaudal progression of development, where infants learn to use their upper limbs before their lower limbs. However, this relationship may be altered if PA patterns were examined in smaller age groups. Examinations of daily PA patterns and subsequent correlates of PA within specific groups on the basis of age (eg, aged 6 months) or behavior (eg, pre- vs postcrawling infants) are both warranted.
This study has several limitations. The small sample size limits our statistical power in analyzing PA. Future research should consider recruiting a larger sample of children with and without DS, as this would allow for additional analysis (eg, splitting each group to examine pre- and postcrawling activity differences). Although this study is strengthened by the use of accelerometry to measure PA, the lack of biological meaning for this data limits the interpretation of results. Cut points specific to the infant age group are in need.20,22 Furthermore, a recent study developed an algorithm to identify infant leg movements from other background movement using analyses of linear and rotational acceleration with a wearable inertial sensor.21 Although the accelerometers used in the present study are not as advanced, further methodological innovations are needed to effectively measure movement, remove sources of external motion, and understand PA during infancy. A 30-minute interval was selected as our recording increment on the parental log. Although parents were encouraged to select a categorization of PA, which represented the majority of time spent in any given activity, it is possible that this method could result in classification errors. For example, in infants younger than 6 months, a considerable amount of parental handling (ie, carrying from room to room) may occur within 30 minutes but not represent the majority of activity time during the respective interval. Future researchers should examine the use of a shorter recording interval to more accurately describe the PA patterns during each interval. Fortunately, this study clearly demonstrates that most parents are able to complete an activity log that is not too complex or burdensome.
This study provides objective PA data on infants with DS and those developing typically. The knowledge gained in this study can be used in establishing PA patterns and differences in 2 different infant populations. Furthermore, a greater understanding of PA patterns early on in development will assist service providers in determining the efficacy of PA interventions.
1. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA. 2014;311:806–814.
2. Foerste T, Sabin M, Reid S, Reddihough D. Understanding the causes of obesity in children with trisomy 21: hyperphagia vs physical inactivity. J Intellect Disabil Res. 2016;60:856–864.
3. Blake-Lamb TL, Locks LM, Perkins ME, Baidal JAW, Cheng ER, Taveras EM. Interventions for childhood obesity in the first 1,000 days: a systematic review. Am J Prev Med. 2016;50:780–789.
4. Laws R, Campbell KJ, van der Pligt P, et al The impact of interventions to prevent obesity or improve obesity related behaviours in children (0-5 years) from socioeconomically disadvantaged and/or indigenous families: a systematic review. BMC Public Health. 2014;14:779.
5. Llewellyn A, Simmonds M, Owen CG, Woolacott N. Childhood obesity as a predictor of morbidity in adulthood: a systematic review and meta-analysis. Obes Rev. 2016;17:56–67.
6. Bertapelli F, Pitetti K, Agiovlasitis S, Guerra-Junior G. Overweight and obesity in children and adolescents with down syndrome - prevalence, determinants, consequences, and interventions: a literature review. Res Dev Disabil. 2016;57:181–192.
7. Myers A, Gibbons C, Finlayson G, Blundell J. Associations among sedentary and active behaviours, body fat and appetite dysregulation: investigating the myth of physical inactivity and obesity. Br J Sports Med. 2016;0:1–6. doi:10.1136/bjsports-2015-095640.
8. World Health Organization. International Classification of Functioning, Disability and Health: ICF. Geneva, Switzerland: World Health Organization; 2001.
9. Barr M, Shields N. Identifying the barriers and facilitators to participation in physical activity
for children with Down syndrome. J Intellect Disabil Res. 2011;55:1020–1033.
10. Whitt-Glover MC, O'Neill KL, Stettler N. Physical activity
patterns in children with and without Down syndrome. Pediatr Rehabil. 2006;9:158–164.
11. Esposito PE, MacDonald M, Hornyak JE, Ulrich DA. Physical activity
patterns of youth with Down syndrome. Intellect Dev Disabil. 2012;50:109–119.
12. Raiten DJ, Raghavan R, Porter A, Obbagy JE, Spahn JM. Executive summary: evaluating the evidence base to support the inclusion of infants
and children from birth to 24 mo of age in the Dietary Guidelines for Americans—“the B-24 Project.” Am J Clin Nutr. 2014;99:663–691.
13. Wojcicki JM, Heyman MB. Let's move—childhood obesity prevention from pregnancy and infancy onward. N Engl J Med. 2010;362:1457–1459.
14. SHAPE America. Active Start: A Statement of Physical Activity
Guidelines for Children Birth-Age 5. Reston, VA: Society of Health and Physical Educators; 2009.
15. Tremblay MS, LeBlanc AG, Carson V, et al Canadian physical activity
guidelines for the early years (aged 0-4 years). Appl Physiol Nutr Metab. 2012;37:345–356.
16. Wentz E. The Impact of an Early and Intense Prone Positioning Program in Infants
With and Without Down Syndrome [dissertation]. Ann Arbor, MI: University of Michigan School of Kinesiology.
17. Angulo-Barroso R, Burghardt AR, Lloyd M, Ulrich DA. Physical activity
with Down syndrome receiving a treadmill intervention. Infant Behav Dev. 2008;31:255–269.
18. Ulrich DA, Lloyd MC, Tiernan CW, Looper JE, Angulo-Barroso RM. Effects of intensity of treadmill training on developmental outcomes and stepping in infants
with Down syndrome: a randomized trial. Phys Ther. 2008;88:114–122.
19. Ulrich DA, Ulrich BD, Angulo-Kinzler RM, Yun J. Treadmill training of infants
with Down syndrome: evidence-based developmental outcomes. Pediatrics. 2001;108:e84.
20. Timmons BW, LeBlanc AG, Carson V, et al Systematic review of physical activity
and health in the early years (aged 0-4 years). Appl Physiol Nutr Metab. 2012;37:773–792.
21. Smith B, Trujillo-Priego I, Lane C, Finley J, Horak F. Daily quantity of infant leg movement: wearable sensor algorithm and relationship to walking onset. Sensors. 2015;15:19006–19020.
22. Prioreschi A, Micklesfield LK. A scoping review examining physical activity
measurement and levels in the first 2 years of life. Child Care Health Dev. 2016;42:775–783.
23. Bayley N. Bayley Scales of Infant and Toddler Development. 3rd ed. San Antonio, TX: Harcourt Assessment; 2006.
24. Cole T, Henson G, Tremble J, Colley N. Birthweight for length: Ponderal index, body mass index or Benn index? Ann Hum Biol. 1997;24:289–298.
25. World Health Organization. WHO Child Growth Standards: Length/Height for Age, Weight-for-Age, Weight-for-Length, Weight-for-Height and Body Mass Index-for-Age, Methods and Development. Geneva, Switzerland: World Health Organization; 2006.
26. Cliff DP, Reilly JJ, Okely AD. Methodological considerations in using accelerometers
to assess habitual physical activity
in children aged 0-5 years. J Sci Med Sport. 2009;12:557–567.
27. Worobey J, Vetrini NR, Rozo EM. Mechanical measurement of infant activity: a cautionary note. Infant Behav Dev. 2009;32:167–172.
28. Pitetti K, Baynard T, Agiovlasitis S. Children and adolescents with Down syndrome, physical fitness and physical activity
. J Sport Health Sci. 2013;2:47–57.