Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Jun 1.
Published in final edited form as: Med Sci Sports Exerc. 2010 Jun;42(6):1072–1078. doi: 10.1249/MSS.0b013e3181c619b2

Early Physical Activity Provides Sustained Bone Health Benefits Later in Childhood

Kathleen F Janz 1,2, Elena M Letuchy 2, Julie M Eichenberger Gilmore 3, Trudy L Burns 2,4, James C Torner 2, Marcia C Willing 4, Steven M Levy 3,2
PMCID: PMC2874089  NIHMSID: NIHMS166916  PMID: 19997029

Abstract

Purpose

This study examined the potential effect of early childhood moderate and vigorous physical activity (MVPA) on later bone health.

Methods

Three hundred and thirty-three children, participating in the Iowa Bone Development Study, were studied at ages 5, 8, and 11. MVPA (min/d) was measured using an accelerometry-based physical activity monitor. Bone mineral content (BMC, g) of the whole body, lumbar spine, and hip was measured using dual energy x-ray absorptiometry (DXA). Mixed regression models were used to test whether age 5 MVPA had an effect on BMC at ages 8 and 11, after adjustment for concurrent height, weight, age, maturity, and MVPA. The analysis was repeated to control for age 5 bone outcomes. Mixed model least-squares means at the person level of covariates for age group were used to compare the BMC at age 8 and 11 of children in the highest and lowest quartiles of age 5 MVPA.

Results

For boys and girls, age 5 MVPA predicted BMC adjusted for concurrent height, weight, age, maturity, and MVPA at ages 8 and 11 (p<0.05). When the analysis was repeated to also control for age 5 BMC, the effect of age 5 MVPA was significant for boys but not girls. Boys and girls in the highest age 5 MVPA quartile had 4 to 14% more BMC at age 8 and 11 than those in the lowest age 5 MVPA quartile (p < 0.05).

Conclusion

These results provide support for the benefits of early MVPA on sustained bone health during childhood especially for boys. Results indicate the importance of increasing MVPA as a strategy to improve BMC later in childhood.

Keywords: accelerometry, bone mineral content. children, DXA, longitudinal, skeletal

INTRODUCTION

Physical activity during childhood has been shown to have immediate bone health benefits. Results from cross-sectional, longitudinal, and randomized intervention studies indicate higher bone mineral content (BMC) in active children when compared to less active peers (1, 9; 16, 17). If higher BMC is maintained into early adulthood, an increase in peak bone mass would be expected; and peak bone mass is thought to be a critical factor in the reduction in fracture risk during middle and late adulthood (11). However, at this time it is not clear if the benefit in BMC accrual associated with early physical activity is sustained if physical activity is reduced during adolescence. This is an important public health issue given the known decrease in physical activity associated with aging, particularly its precipitous drop during adolescence and continued decline during adulthood (7, 27).

Several carefully controlled animal studies indicate that targeted mechanical loading of bone in young animals results in lifetime structural changes (21, 25, 31). Retrospective studies of athletes (gymnasts, tennis players, and ballet dancers) suggest sustained benefits of early activity to the mature skeleton, including greater BMC (2, 13, 14). Other studies, however, suggest that the cessation of sport participation is associated with a reduction in BMC (19, 30). Recently, Gunter and colleagues (6) showed sustained effects from a short-term intervention study of peri-pubertal non-athletes (n = 57, ages 7 to 8 at baseline). In this study, children completing high-impact jumping exercises had 3.6% greater hip BMC than controls at the completion of the 7-month intervention and 1.4% greater hip BMC after approximately 7 years with no additional intervention.

Although studies of athletes and children engaged in targeted interventions provide unique and important information on the potential impact of exercise on BMC accrual, the mechanical loads incurred during these activities may not be conducive to public health recommendations; therefore, population-based studies are needed to set realistic recommendations for the amounts and types of physical activity needed to ensure bone health benefits. Ideally, population-based studies should be prospective to avoid the known limitations of retrospective studies including bias in the measurement of the physical activity exposure. Using a prospective longitudinal study design, Baxter-Jones and colleagues (3) recently reported that physically active children and adolescents (n = 154, ages 8 to 15 at baseline) had 8 to 10% more hip BMC in young adulthood (ages 23 to 30) when compared to less active peers, even after controlling for their adult physical activity levels. This study provides intriguing evidence of the long-term benefits of childhood physical activity on adolescent and adult BMC and supports conjecture that the promotion of physical activity in children could help to prevent osteoporosis later in life. In this report, we extend the work of Baxter-Jones et al. (3) by examining the effect of an early physical activity exposure (age 5) on BMC at age 8 and 11 in a cohort of normal healthy children (n = 333). We hypothesized that greater levels of early childhood physical activity would provide a sustained benefit in BMC evident 3 to 6 years later.

METHODS

Participants

Participants were part of the Iowa Bone Development Study, which is a longitudinal study of bone health during childhood and has been described elsewhere (8). Four hundred thirty-three cohort children participated in physical activity and BMC measurements at approximately 5 years of age (baseline), 495 at approximately 8 years of age, and 406 at approximately 11 years of age. Children with data from all three visits (n = 333, 185 boys and 148 girls) were included in the analyses. Mean age was 5.3 ± 0.4 yr at baseline, 8.8 + 0.6 at 3-year follow-up, and 11.2 + 0.3 yr at 6-year follow-up. Mean baseline heights and weights did not differ significantly (p > 0.05) between children with physical activity and BMC data at all three visits and those without full data. However, there was a small mean baseline age difference between children with data at all three visits and those without (5.3 yr vs. 5.2 yr; p < 0.05). The parents of the children provided written informed consent and the children provided assent. The University of Iowa Institutional Review Board approved this study.

Physical Activity

Physical activity (movement counts) was measured using the ActiGraph uniaxial physical activity monitor (model number 7164, Pensacola, FL). Procedures for physical activity measurement using the ActiGraph and validation of this monitor have been described elsewhere (4, 10, 28). Children were asked to wear the monitors all day during waking hours for 4 consecutive days at ages 5 and 8. The number of wear days was increased at age 11 to 5 days to account for increased day-to-day variability in accelerometry-measured physical activity in older children when compared to younger children (10). To reduce seasonal effects, physical activity was only monitored during the autumn (September through November). Monitors were distributed to include at least one weekend day at age 5, 8, and 11. Children were included in the analysis of data if they wore the accelerometer at least 8 hours/day for at least 3 days and within 15 months of the dual energy x-ray absorptiometry (DXA) scan. One-minute epochs were used to sum movement count values and a summary variable of daily minutes spent in moderate and vigorous physical activity (MVPA) was calculated. The cut-point threshold of 3000 accelerometer movement counts per minute (ct•min−1) was used to define MVPA. This cut-point has been associated with MVPA at normal walking speeds in children (5).

Bone Mineral Content

At ages 5 and 8, whole body, anterior-posterior (AP) lumbar spine, and left hip scans were obtained using a Hologic QDR 2000 DXA (Hologic, Inc., Bedford, MA) with software version 7.20B. The fan-beam mode was used for spine and whole body scans and the pencil beam mode was used for the hip scan. At age 11, the Hologic QDR 4500 DXA (Delphi upgrade) with software version 12.3 and fan-beam mode was used for all scan acquisitions. All scans (ages 5, 8, and 11) were re-analyzed using Hologic software version 12.6. BMC (g) was derived from the scan images. Previous research suggests that skull size confounds whole body bone data in young children (26); therefore, all whole body results presented in this study represent BMC excluding the skull. Software-specific Global Regions of Interest (ROI) were used to designate the general boundaries of the hip and spine images. A review of the bone within the ROI box was confirmed by the operator and edited to ensure appropriate bone-edge detection. Quality control scans were performed daily using the Hologic spine phantom. To minimize operator-related variability, all measurements were conducted by one of three experienced technicians. The precision error for BMC measurements is low in our laboratory (coefficient of variation of < 1% for quality control scans performed daily using the Hologic phantom). Translational equations from 4500 DXA measures to 2000 DXA measures for age 11 records were used to adjust for the differences between the two DXA machines. To develop these translational equations a separate study was conducted where 60 of the children (32 boys, 28 girls) ages 9.9 to 12.4 (mean 11.4, SD 0.4) were scanned on each machine in random order during one clinic visit. The actual observations were closely aligned around the translational equation regression line and the coefficient for determination (R2) for the 4500 DXA data regressed on to the 2000 DXA data was 0.99 (unpublished observations).

Anthropometry and Somatic Maturity

Research-trained nurses measured the child’s height (cm) using a Harpenden stadiometer (Holtain, UK) and body mass (kg) using a Healthometer physician’s scale (Continental, Bridgeview, IL) at each visit (age 5, 8, and 11 yr). At age 11, sitting height was measured and used to estimate maturity offset (year from peak height velocity) using predictive equations established by Mirwald and colleagues (20). This approach includes age, gender, weight, height, sitting height, and leg length as predictors of years from peak height velocity (or somatic maturity). The Mirwald method has been validated in white Canadian children and adolescents (R2=0.91–0.92, SEE=0.49–0.50) (20). We dichotomized the maturity offset variable as 0 (prior to peak height velocity or pre-mature) or 1 (≥ peak height velocity or mature).

Statistical Analysis

Gender-specific descriptive analyses including Student’s t-tests were conducted of measures obtained at age 5, 8, and 11. Mixed regression models for correlated data were used to examine whether physical activity at age 5 predicted BMC at ages 8 and 11 with adjustment for concurrent (at age 8 or 11) height, weight, age, somatic maturity, and MVPA. The residual observations within children were correlated through the within-person variance–covariance matrix and the appropriate matrix structure was determined based on Akaike’s Information Criterion (AIC) for goodness of fit. An unstructured variance-covariance matrix was selected to allow for the higher variance of age 11 measures, together with the within-person covariance. Residual and studentized residual graphs were used to confirm the models’ assumptions and fit (not shown). The analysis was repeated to also control for BMC (whole body, spine, or hip) at age 5.

To investigate the effect of high versus low age 5 physical activity, the high (most active) and the low (least active) quartile groups based on age 5 MVPA were identified for the whole sample and the analysis was stratified by gender. The highest quartile group consisted of 37 girls and 47 boys; whereas, lowest quartile group consisted of 25 boys and 59 girls. Mixed model least-squares means at the person level of covariates for age group were used to compare the BMC at age 8 and 11 of children in the highest and lowest quartiles of age 5 MVPA. The covariates included concurrent height, weight, age, somatic maturity, and MVPA. Least-squares BMC means for children at age 5 were calculated in a separate cross-sectional model. The approach was then repeated to include BMC at age 5 as a covariate in the age 8 and 11 group analyses. The level of significance was set at 0.05 for all analyses which were conducted using SAS (version 9.1.3).

RESULTS

Participant Characteristics

The characteristics of children at the time of each examination (age 5, 8, and 11), including age, height, weight, BMC, and MVPA are summarized in Table 1. Throughout the study, boys on average engaged in more MVPA than did girls (p < 0.05) and this difference increased as the children aged. Gender-related differences were also noted in BMC. At ages 5 and 8, the boys had greater spine BMC, and at age 8 the boys had greater hip BMC than girls. At age 11, the girls had greater spine BMC than boys.

Table 1.

Characteristics of the participants (n = 185 girls and 148 boys)

Age 5 Age 8 Age 11
Girls Boys Girls Boys Girls Boys
Mean ± SD Mean ± SD Mean ± SD Mean ± SD Mean ± SD Mean ± SD
Ht (cm) 111.1 ± 5.5 112.0 ± 5.5 132. 5 ± 6.8* 134.1 ± 7.4 149.1 ± 7.7 148.9 ± 7.7
Wt (kg) 20.0 ± 3.8 20.3 ± 3.4 31.7 ± 8.7 32.2 ± 8.5 44.6 ± 12.2 44.1 ± 12.4
Age at DXA (yr) 5.3 ± 0.4 5.2 ± 0.4 8.7 ± 0.6 8.7 ± 0.6 11.2 ± 0.3 11.2 ± 0.3
Whole Body BMC (g) 247.5 ± 80.6 246.2 ± 75.8 560.3 ± 165.8 581.5 ± 252.9 951.3 ± 267.1. 918.0 ± 253.0
Spine BMC (g) 14.0 ± 2.6* 14.6 ± 2.6 21.8 ± 4.4* 22.9 ± 4.5 31.9 ± 8.4* 29.6 ± 5.8
Hip BMC (g) 5.9 ± 1.3 6.1 ± 1.4 11.2 ± 2.6* 11.9 ± 3.0 17.7 ± 4.2 18.0 ± 4.4
MVPA (min/d) 24.2 ± 13.4* 31.1 ± 16.3 25.2 ± 13.1* 40.6 ± 21.4 22.6 ± 13.6* 42.3 ± 22.7
Open in a new tab

Note: No children were classified as mature at age 5 or age 8. At age 11, 19% of girls and no boys were classified as mature.

*

p < 0.05 Student’s t-test of girls vs. boys. BMC = bone mineral content, MVPA = moderate and vigorous physical activity.

Effect of Early Physical Activity on Later BMC

Gender-specific regression models for BMC are presented in the left half of Tables 2 and 3 in Panel A. After adjustment for concurrent (age 8 or 11) age, height, weight, somatic maturity, and MVPA, age 5 MVPA was a significant predictor of age 8 and 11 BMC in both boys and girls at the whole body, spine, and hip (p < 0.05) (Table 2 and 3, Panel A). The age 5 MVPA β (slopes) indicated that the contribution of age 5 MVPA to age 8 and 11 BMC was consistently greater in boys than girls. In all models for girls and in the whole body and spine models for boys, concurrent (age 8 or 11) MVPA was not significantly (p > 0.05) associated with BMC at age 8 and 11 when age 5 MVPA was included in the models. For boys, concurrent MVPA remained significantly associated with hip BMC at age 8 and 11 (p = 0.004) after age 5 MVPA entered the model.

Table 2.

Mixed regression model analysis of BMC for girls (n = 185) at ages 8 and 11 as predicted by MVPA at age 5 years

Panel A without Age 5 BMC in the Model Panel B with Age 5 BMC in the Model
Effect β SE p-value Effect β SE p-value
WB BMC Intercept −487.80 272.91 0.076 Intercept 48.23 240.47 0.841
Age (yr) −165.02 52.57 0.002 Age (yr) −202.06 46.70 <.001
Age2 (yr2) 9.36 2.74 0.001 Age2 (yr2) 13.14 2.47 <.001
Maturity (0,1) 84.83 19.57 <.001 Maturity (0,1) 119.49 20.60 <.001
Ht (cm) 10.48 0.91 <.0001 Ht (cm) 6.10 0.80 <.001
Wt (kg) 11.39 0.67 <.001 Wt (kg) 8.84 0.63 <.001
Concurrent MVPA (10 min/d) 0.97 2.83 0.732 Concurrent MVPA (10 min/d) 3.55 2.40 0.142
Age 5 MVPA (10 min/d) 7.40 3.51 0.037 Age 5 MVPA (10 min/d) 1.52 2.84 0.594
Age 5 WB BMC (g) 0.69 0.07 <.001
Spine BMC Intercept −8.49 11.14 0.447 Intercept 9.60 9.16 0.296
Age (yr) −6.24 2.10 0.003 Age (yr) −6.79 1.79 0.002
Age2 (yr2) 0.32 0.11 0.003 Age2 (yr2) 0.44 0.09 <.001
Maturity (0,1) 5.24 0.71 <.001 Maturity (0,1) 6.28 0.77 <.001
Ht (cm) 0.41 0.04 <.001 Ht (cm) 0.15 0.03 <.001
Wt (kg) 0.13 0.03 <.001 Wt (kg) 0.09 0.02 <.001
Concurrent MVPA (10 min/d) 0.02 0.12 0.871 Concurrent MVPA (10 min/d) 0.14 0.09 0.133
Age 5 MVPA (10 min/d) 0.39 0.17 0.022 Age 5 MVPA (10 min/d) 0.06 0.11 0.594
Age 5 Spine BMC (g) 1.03 0.07 <.001
Hip BMC Intercept −15.75 4.88 0.002 Intercept −7.32 4.54 0.108
Age (yr) −2.02 0.91 0.029 Age (yr) −2.53 0.86 0.004
Age2 (yr2) 0.12 0.05 0.01 Age2 (yr2) 0.18 0.04 <.001
Maturity (0,1) 1.91 0.30 <.001 Maturity (0,1) 2.38 0.32 <.001
Ht (cm) 0.24 0.02 <.001 Ht (cm) 0.15 0.02 <.001
Wt (kg) 0.10 0.02 <.001 Wt (kg) 0.07 0.01 <.001
Concurrent MVPA (10 min/d) 0.08 0.05 0.117 Concurrent MVPA (10 min/d) 0.08 0.05 0.076
Age 5 MVPA (10 min/d) 0.18 0.08 0.026 Age 5 MVPA (10 min/d) 0.04 0.06 0.551
Age 5 Hip BMC (g) 0.78 0.08 <.001
Open in a new tab

Note: β = regression parameter estimate, WB BMC = whole body, MVPA = moderate and vigorous physical activity in ten minutes intervals to avoid small coefficients. Unstructured within person error-covariance was used for repeated measures, variance-covariance values were highly significant (p < 0.001). Akaike’s Information Criterion (AIC) values indicated that this strategy improved the model fit when compared to a compound symmetric structure.

Table 3.

Mixed regression model analysis of BMC for boys (n = 148) at age 8 and age 11 as predicted by MVPA at age 5 years

Panel A without Age 5 BMC in the Model Panel B with Age 5 BMC in the Model
Effect β SE p-value Effect β SE p-value
WB BMC Intercept −608.93 258.70 0.020 Intercept −176.01 248.36 0.48
Age (yr) −82.50 51.87 0.114 Age (yr) −134.44 49.56 0.008
Age2 (yr2) 5.69 2.64 0.033 Age2 (yr2) 9.68 2.56 0.002
Ht (cm) 7.78 1.07 <.001 Ht (cm) 4.96 0.92 <.001
Wt (kg) 12.44 0.81 <.001 Wt (kg) 10.08 0.75 <.001
Concurrent MVPA (10 min/d) 0.57 2.07 0.784 Concurrent MVPA (10 min/d) 1.06 1.87 0.574
Age 5 MVPA (10 min/d) 8.95 3.40 0.009 Age 5 MVPA (10 min/d) 5.77 2.87 0.046
Age 5 WB BMC (g) 0.72 0.08 <.001
Spine BMC Intercept −15.69 7.03 0.027 Intercept −8.03 7.21 0.267
Age (yr) −2.87 1.38 0.04 Age (yr) −2.60 1.43 0.072
Age2 (yr2) 0.15 0.07 0.031 Age2 (yr2) 0.19 0.07 0.009
Ht (cm) 0.34 0.04 <.001 Ht (cm) 0.16 0.04 <.001
Wt (kg) 0.12 0.02 <.001 Wt (kg) 0.10 0.02 <.001
Concurrent MVPA (10 min/d) 0.02 0.06 0.779 Concurrent MVPA (10 min/d) 0.04 0.06 0.542
Age 5 MVPA (10 min/d) 0.75 0.14 <.001 Age 5 MVPA (min/d) 0.44 0.11 0.002
Age 5 Spine BMC (g) 0.8838 0.08 <.001
Hip BMC Intercept −9.43 4.35 0.032 Intercept −2.03 4.43 0.647
Age (yr) −2.00 0.86 0.023 Age (yr) −2.58 0.88 0.004
Age2 (yr2) 0.14 0.04 0.003 Age2 (yr2) 0.19 0.04 <.001
Ht (cm) 0.18 0.02 <.001 Ht (cm) 0.09 0.02 <.001
Wt (kg) 0.13 0.02 <.001 Wt (kg) 0.12 0.01 <.001
Concurrent MVPA (10 min/d) 0.11 0.04 0.004 Concurrent MVPA (10 min/d) 0.12 0.03 0.005
Age 5 MVPA (10 min/d) 0.26 0.07 0.004 Age 5 MVPA (min/d) 0.13 0.06 0.039
Age 5 Hip BMC (g) 0.81 0.08 <.001
Open in a new tab

Note: β = regression parameter estimate, WB BMC = whole body, MVPA = moderate and vigorous physical activity in ten minutes intervals to avoid small coefficients. The models for boys were not adjusted for maturity, because no boys were categorized as mature. Unstructured within person error-covariance was used for repeated measures, variance-covariance values were highly significant (p<0.001). Akaike’s Information Criterion (AIC) values indicated that this strategy improved the model fit when compared to a compound symmetric structure.

To examine the relationship between age 5 MVPA and BMC at age 8 and 11, independent of the age 5 MVPA contribution to age 5 BMC (which presumably contributed to age 8 and 11 BMC), regression models were fitted that included age 5 BMC (see right sides of Tables 2 and 3, Panel B) in addition to concurrent (age 8 or 11) age, height, weight, somatic maturity, and MVPA. This is a conservative approach that would be expected to attenuate the associations since the immediate benefit of age 5 MVPA on age 5 BMC is controlled. Therefore, what is tested is the potential impact beyond the age 5 time period. Age 5 MVPA remained a significant predictor of age 8 and 11 whole body BMC, spine BMC, and hip BMC in boys, but not girls. In these models with the exception of hip BMC in boys, concurrent (age 8 or age 11) MVPA was not significantly associated with BMC (p > 0.05).

Differences in BMC at Age 8 and 11 between the Least and Most Active Age 5 Children

A summary of characteristics of the children who were in the lowest (least active) and highest (most active) MVPA quartiles at age 5 is presented in Table 4. At age 5 there was no difference in age, weight, or height between the boys in the lowest and highest MVPA quartiles. At age 5, girls in the lowest MVPA quartile were heavier than girls in the highest MVPA quartile (p < 0.05). However, the absolute difference in weight between girls in the lowest and highest age 5 MVPA quartile was 1.16 kg, which was less than the absolute difference in weight between boys in the lowest and highest age 5 MVPA quartile. These data suggest that the statistically significance difference in weight in girls was due to our larger sample of girls (when compared to boys).

Table 4.

Characteristics of children in the lowest (n=84) and highest quartiles (n=84) of age 5 MVPA

Age 5 Age 8 Age 11
Lowest Highest Lowest Highest Lowest Highest
Mean ± SD Mean ± SD Mean ± SD Mean ± SD Mean ± SD Mean ± SD
Girls
Ht (cm) 111.4 ± 6.2 111.0 ± 5.0 133.2 ± 7.4 131.8 ± 6.0 149.2 ± 8.6 149.0 ± 6.4
Wt (kg) 20.9 *± 5.1 19.7 ± 2.5 33.6 * ± 11.3 30.0 ± 6.2 45.8 ± 14.5 43.6 ± 9.6
Age at DXA (yr) 5.3 ± 0.5 5.3 ± 0.4 8.7 ± 0.6 8.6 ± 0.7 11.3 ± 0.3 11.2 ± 0.3
MVPA (min/d) 11.2 * ± 3.5 45.4 ± 10.0 19.4 * ± 10.7 34.1 ± 13.7 19.3 * ± 10.8 30.4 ± 17.7
Boys
Ht (cm) 111.8 ± 7.3 112.6 ± 5.4 134.3 ± 9.2 134.8 ± 6.9 148.6 ± 9.3 149.8 ± 7.5
Wt (kg) 21.5 ± 5.4 20.1 ± 2.5 34.6 ± 12.1 31.8 ± 7.0 47.1 ± 16.4 43.8 ± 10.4
Age at DXA (yr) 5.2 ± 0.4 5.3 ± 0.4 8.6 ± 0.6 8.7 ± 0.6 11.1 * ± 0.3 11.3 ± 0.3
MVPA (min/d) 9.55 * ± 3.6 50.7 ± 11.0 20.4 * ± 14.4 50.5 ± 22.2 29.4 * ± 16.9 53.5 ± 26.0
Open in a new tab

Note: 59 girls and 25 boys in lowest age 5 MVPA quartile; 37 girls and 47 boys in highest age 5 MVPA quartile.

*

p < 0.05 for lowest vs. highest age 5 MVPA.

Least-squares BMC means for the lowest and the highest quartiles of age 5 MVPA are presented in Table 5. After adjustment for concurrent age, height, weight, and MVPA, boys and girls in the highest age 5 MVPA quartile had significantly greater BMC at age 8 and 11 than those in the lowest quartile (p < 0.05). On average, boys in the highest age 5 MVPA quartile had 8, 14, and 11% more BMC at the whole body, spine, and hip (respectively) at age 8 than boys in the lowest age 5 MVPA quartile. Girls in the highest age 5 MVPA quartile had 6, 8, and 8% more BMC at the whole body, spine, and hip (respectively) at age 8 than girls in the lowest age 5 MVPA quartile. At age 11, the difference between the highest and lowest quartile of age 5 MVPA in boys decreased to 5% for the whole body, 7% for the spine, and 7% for the hip. The age 11 BMC difference in girls was 4% for the whole body, 6% for the spine, and 5% for the hip. After controlling for age 5 BMC, BMC differences at age 8 and age 11 for girls were not significant (p > 0.05) (see table 6). However for boys, BMC differences at age 8 and age 11 remained significant at the spine and hip. On average, the BMC difference in boys for the spine was 7% at age 8 and 5% at age 11. For the hip, the BMC difference in boys was 6% at age 8 and 4% at age 11.

Table 5.

Mixed-model least-squares means for BMC in children in the lowest and highest quartiles of MVPA at age 5 years

Age 5 Age 8 Age 11
Age 5 Lowest Age 5 Highest Age 5 Lowest Age 5 Highest Age 5 Lowest Age 5 Highest
LS Mean(SE) LS Mean(SE) LS Mean(SE) LS Mean(SE) LS Mean(SE) LS Mean(SE)
Boys
WB BMC 225.31 (7.29) 259.93 (5.30)*** 551.15 (13.56) 593.98 (9.16)* 894.6 (15.99) 937.43 (12.94)*
Spine BMC 13.62 (0.32) 15.49 (0.23)*** 21.41 (0.59) 24.37 (0.43)*** 29.21 (0.39) 31.15 (0.46)***
Hip BMC 5.41 (0.18) 6.52 (0.13)*** 11.21 (0.30) 12.42 (0.21)** 17.29 (0.34) 18.50 (0.27)**
Girls
WB BMC 236.11 (4.44) 259.05 (5.55)** 541.23 (7.85) 574.77 (10.06)* 930.78 (11.09) 964.33 (12.17)*
Spine BMC 13.64 (0.25) 14.48 (0.31)* 21.11 (0.38) 22.86 (0.49)** 31.15 (0.49) 32.90(0.58)**
Hip BMC 5.63 (0.11) 6.22 (0.13)*** 10.82 (0.18) 11.63 (0.23)** 17.39 (0.22) 18.20 (0.26)**
Open in a new tab
***

p < 0.001,

**

p < 0.01,

*

p < 0.05

Note: WB BMC = whole body BMC. BMC at age 5 years was adjusted for concurrent (age 5 years) age, height, and weight. BMC at age 8 years and age 11 years was adjusted for concurrent age, height, weight, maturity, and MVPA (moderate and vigorous physical activity)

Table 6.

Mixed-model least-squares means for BMC in children in the lowest and highest quartiles of MVPA at age 5 years, with adjustment for age 5 BMC

Age 8 Age 11
Age 5 Lowest Age 5 Highest Age 5 Lowest Age 5 Highest
LS Mean(SE) LS Mean(SE) LS Mean(SE) LS Mean(SE)
Boys
WB BMC 557.14 (11.05) 584.83 (7.97) 901.74 (11.16) 929.43 (11.33)
Spine BMC 22.16 (0.45) 23.62 (0.32)* 28.95 (0.48) 30.41 (0.37)*
Hip BMC 11.49 (0.24) 12.12 (0.17)* 17.57 (0.28) 18.20 (0.23)*
Girls
WB BMC 544.68 (6.17) 560.83 (8.01) 932.89 (10.1 9) 949.03 (11.43)
Spine BMC 21.37 (0.24) 22.14 (0.32) 31.36 (0.40) 33.13 (0.44)
Hip BMC 10.93 (0.14) 11.30 (0.18) 17.50 (0.19) 17.86 (0.22)
Open in a new tab
*

p-value<0.05

Note: WB BMC = whole body BMC. . BMC at age 8 years and age 11 years was adjusted for concurrent age, height, weight, maturity, MVPA (moderate and vigorous physical activity), and age 5 BMC.

DISCUSSION

Little is known about habitual physical activity and BMC accrual in children younger than 8 yr (3). This is partially due to the challenges of measuring “everyday” physical activity in young children. To the best of our knowledge, the Iowa Bone Development Study is the first study to investigate longitudinal associations between objectively-measured physical activity and BMC in young children (8). Of the currently available objective instruments for the assessment of children’s physical activity, accelerometry-based physical activity monitors, such as the ActiGraphs used in our study, appear to be the most appropriate choice when bone outcomes are of interest. These monitors detect acceleration (m/s2) which is proportional to the muscular and impact load forces acting on the skeletal system (22). In addition, they provide an absolute measure that is, for the most part, independent of fitness, adiposity, and size; factors that are likely to change during the course of a longitudinal study.

In general, pediatric researchers have focused on children in middle and late childhood with the assumption that the time period just prior to puberty represents a “window of opportunity” when the skeleton is most sensitive to mechanical loading (12, 23). Results from our work suggest widening the “window of opportunity” concept to include an earlier time period (age 5). Intervening early to increase physical activity may have the added advantage of improving compliance, since most young children are amenable, and even eager, to engage in age-appropriate physical activities and they have the discretionary time to do so.

To illustrate the implications of our findings, we calculated the average change in hip BMC expected with a 30 minute increase in MVPA at age 5. Thirty minutes was selected given the mean MVPA at age 5 in our cohort was approximately 28 min and current U.S. federal guideline for children is 60 minutes of daily age-appropriate moderate and vigorous activity (29). The mixed regression results for girls (presented in Table 2-Panel A) suggest that, if all other factors were held constant, 30 minutes of additional MVPA at age 5 years would result, on average, in an additional 0.5 g of hip BMC later in childhood. In our girls’ cohort, 0.5 g was 4.5% of the mean hip BMC at age 8 and 3.0% at age 11. Using this same approach, mixed regression results for boys (Table 3-Panel A) suggest 30 minutes of additional MVPA at age 5 years would result, on average, in an additional 0.8 g of hip BMC later in childhood. The additional hip BMC (0.8 g) was 6.7% of the mean hip BMC of our boys’ cohort at age 8 and 4.4% at age 11. Increases of this magnitude are similar to what is expected with targeted loading interventions (6, 16, 17) and likely to provide a meaningful increase in peak bone mass later in childhood (12).

We have consistently noted stronger associations between physical activity and BMC in boys (when compared to girls) in our cohort (8, 9). McKay and colleagues (18) have also reported stronger relationships between physical activity and BMC in boys than girls. These findings are consistent with McDonald and colleagues suggestion of a sex-specific sensitivity of bone to mechanical loading that favors males (15). However, we also suspect that the stronger relationship observed in boys in our study was at least partially a result of the greater activity at age 5 in boys when compared to girls. On average, the boys in our study were 28% more active than girls at age 5. Our quartile analysis indicated that even within the most active group of age 5 children, the boys were 12% more active than the girls. More activity would result in more mechanical loading of the skeleton and presumably greater benefits.

Limitations of the present study include the use of a convenience sample of mostly Iowa children with low minority representation and relatively high socio-economic status. It is possible that the more active children at age 5 were different from their peers in ways that we did not consider or for which we were unable to control, e.g., dietary intake and genetic factors. Our study results should be confirmed via long-term follow-up of randomized, controlled trials using a more heterogeneous sample. In addition, accelerometry-based physical activity monitoring is considered the preferred method for measuring children’s physical activity in epidemiologic studies. However, our accelerometry method has limitations including our use of a one-minute epoch. This approach may have under-measured MVPA (30). Furthermore, there is no agreed upon age-related cut-point value for ActiGraph accelerometry measures for longitudinal studies. However, our approach is consistent with a 6-year follow-up study used to examine changes in physical activity and metabolic outcomes (7), as well as recent laboratory work suggesting that (accelerometry-derived) movement count values are stable in relationship to oxygen consumption during middle childhood (5).

This report provides evidence of a sustained benefit on BMC associated with early physical activity at the whole body, spine, and hip. Our findings support the hypothesis that there is a pathway between early physical activity and later BMC that is independent of the effect of accumulated physical activity. An important implication of our work is that children who are less physically active at an early age may lose out on the opportunity to obtain the highest peak BMC possible later in life when they are likely to be less active. If the residual benefit observed in this study persists through adolescence, early childhood may prove to be an important developmental period for the promotion of physical activity to optimize peak bone mass during young adulthood and, perhaps, the prevention of osteoporosis during later years.

Acknowledgments

The authors thank the staff of the Iowa Bone Development Study for their organizational efforts. We gratefully acknowledge and thank the children and parents of the Iowa Fluoride Study and the Iowa Bone Development Study, because without their contributions, this work would not have been possible.

The results of the present study do not constitute endorsement by the American College of Sports Medicine (ACSM).

Funding: This study was supported by the National Institute of Dental and Craniofacial Research R01-DE12101 and R01-DE09551, and the General Clinical Research Centers Program from the National Center for Research Resources, M01-RR00059. There are no conflicts of interest to report.

References

  • 1.Bailey DA, McKay HA, Mirwald RL, Crocker PRE, Faulkner RA. A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: The University of Saskatchewan Bone Mineral Accrual Study. J Bone Miner Res. 1999;14:1672–79. doi: 10.1359/jbmr.1999.14.10.1672. [DOI] [PubMed] [Google Scholar]
  • 2.Bass S, Pearch G, Bradney M, Hendrich E, Delmas PD, Harding A, Seeman E. Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepurbertal and retired female gymnasts. J Bone Miner Res. 1998;13:500–507. doi: 10.1359/jbmr.1998.13.3.500. [DOI] [PubMed] [Google Scholar]
  • 3.Baxter-Jones ADG, Kontulaine SA, Faulkner RA, Bailey DA. A longitudinal study of the relationship of physical activity to bone mineral accrual from adolescent to young adulthood. Bone. 2008;43:1101–1107. doi: 10.1016/j.bone.2008.07.245. [DOI] [PubMed] [Google Scholar]
  • 4.Ekelund U, Sjostrom M, Yngve A, et al. Physical activity assessed by activity monitor and doubly labeled water in children. Med Sci Sports Exerc. 2001;33:275–281. doi: 10.1097/00005768-200102000-00017. [DOI] [PubMed] [Google Scholar]
  • 5.Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26:1557–1565. doi: 10.1080/02640410802334196. [DOI] [PubMed] [Google Scholar]
  • 6.Gunter K, Baxter-Jones ADG, Mirwald RL, Almstedt H, Fuchs RK, Durski S, et al. Impact exercise increase BMC during growth: An 8-year longitudinal study. J Bone Miner Res. 2009;23:986–993. doi: 10.1359/JBMR.071201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Jago R, Wedderkopp N, Kristensen PL, Moeller NC, Andersen LB, Cooper AR, Frobert K. Six-year change in youth physical activity and effect on fasting insulin and HOMA-IR. Am J Prev Med. 2008;35:554–560. doi: 10.1016/j.amepre.2008.07.007. [DOI] [PubMed] [Google Scholar]
  • 8.Janz KF, Burns TL, Torner JC, Levy SM, Paulos R, Willing MC, et al. Physical activity and bone measures in young children: The Iowa Bone Development Study. Pediatrics. 2001;107:1387–13931. doi: 10.1542/peds.107.6.1387. [DOI] [PubMed] [Google Scholar]
  • 9.Janz KF, Gilmore JM, Burns TL, Levy SM, Torner JC, Willing MC, et al. Physical activity augments bone mineral accrual in young children: the Iowa Bone Development Study. J Pediatr. 2006;148:793–9. doi: 10.1016/j.jpeds.2006.01.045. [DOI] [PubMed] [Google Scholar]
  • 10.Janz KF, Witt J, Mahoney LT. The stability of children’s physical activity as measured by accelerometry and self-report. Med Sci Sports Exerc. 1995;27:1326–1332. [PubMed] [Google Scholar]
  • 11.Karlsson MK. Does exercise during growth prevent fractures in later life? Med Sport5 Sci. 2007;51:121–36. doi: 10.1159/000103012. [DOI] [PubMed] [Google Scholar]
  • 12.Khan K, McKay HA, Haapasalo H, Bennell KL, Forwood MR, Kannus P, et al. Does childhood and adolescence provide a unique opportunity y for exercise to strengthen the skeleton? J Science Med Sport. 2000;3:150–164. doi: 10.1016/s1440-2440(00)80077-8. [DOI] [PubMed] [Google Scholar]
  • 13.Khan KM, Bennell KL, Hopper JL, Ficker I, Nowson CA, Sherwin AJ, et al. Self reported ballet classes undertaken at age 10–12 years and hip bone mineral density in later life. Osteoporos Int. 1998;8:165–173. doi: 10.1007/BF02672514. [DOI] [PubMed] [Google Scholar]
  • 14.Kontulainen S, Kannus P, Haapasalo H, Sievanen H, Oja P, Vuori I. Changes in bone mineral content with decreased training in competitive young adult tennis players and control: a prospective 4-year follow-up. Med Sci Sports Exerc. 1999;31:646–652. doi: 10.1097/00005768-199905000-00004. [DOI] [PubMed] [Google Scholar]
  • 15.Macdonald HM, Kontulaine S, Petit M, Janssen P, McKay H. Bone strength and its determinants in pre- and early pubertal boys and girls. Bone. 2006;39:598–608. doi: 10.1016/j.bone.2006.02.057. [DOI] [PubMed] [Google Scholar]
  • 16.MacKelvie KJ, McKay HA, Khan K, Crocker PR. A school-based exercise intervention augments bone mineral accrual in early pubertal girls. J Pediatr. 2001;139:501–8. doi: 10.1067/mpd.2001.118190. [DOI] [PubMed] [Google Scholar]
  • 17.MacKelvie KJ, Petit MA, Khan KM, Beck TJ, McKay HA. Bone mass and structure are enhanced following a 2-year randomized controlled trial of exercise in prepubertal boys. Bone. 2004;34:755–764. doi: 10.1016/j.bone.2003.12.017. [DOI] [PubMed] [Google Scholar]
  • 18.McKay HA, Petit MA, Schultz RW, Prior JC, Barr SI, Khan KM. Augmented trochanteric bone mineral density after modified physical education classes: A randomized school-based exercise intervention study in prepubescent and early pubescent children. J of Peds. 2000;136:156–162. doi: 10.1016/s0022-3476(00)70095-3. [DOI] [PubMed] [Google Scholar]
  • 19.Michael BA, Lane NE, Bjorkengren A, Bloch DA, Fries JF. Impact of running on lumbar bone density: a 5-year longitudinal study. J Rheumatol. 1992;19:1759–1763. [PubMed] [Google Scholar]
  • 20.Mirwald RL, Baxter-Jones AD, Bailey DA, et al. An assessment of maturity from anthropometric measurements. Med Sci Sports Exerc. 2002;34:689–694. doi: 10.1097/00005768-200204000-00020. [DOI] [PubMed] [Google Scholar]
  • 21.Modlesky CM, Lewis RD. Does exercise during growth have a long-term effect on bone health? Exerc Sport Sci Rev. 2002;30:171–176. doi: 10.1097/00003677-200210000-00006. [DOI] [PubMed] [Google Scholar]
  • 22.Montoye HJ, Kemper HCG, Saris WHM, Washburn RA. Measuring Physical Activity and Energy Expenditure. Champaign, IL: Human Kinetics; 1996. pp. 72–95. [Google Scholar]
  • 23.Petit MA, McKay HA, MacKelvie KJ, Heinonen A, Khan KM, Beck TJ. A randomized school-based jumping intervention confers site and maturity-specific benefits on bone structural properties in girls: a hip structural analysis study. J Bone Miner Res. 2002;17:363–372. doi: 10.1359/jbmr.2002.17.3.363. [DOI] [PubMed] [Google Scholar]
  • 24.Petit MA, McKay HA, MacKelvie KJ, Heinonen A, Khan KM, Beck TJ. A randomized school-based jumping intervention confers site and maturity-specific benefits on bone structural properties in girls: a hip structural analysis study. J Bone Miner Res. 2002;17:363–372. doi: 10.1359/jbmr.2002.17.3.363. [DOI] [PubMed] [Google Scholar]
  • 25.Silbermann M, Schapira D, Leichter I, Steinberg R. Moderate physical activity through out adulthood increases peak bone mass at middle age and maintains higher trabecular bone density in vertebrae of senescent female rats. Cell Material. 1991;S1:151–158. [Google Scholar]
  • 26.Taylor A, Konrad PT, Norman ME, Harcke HT. Total body bone mineral density in young children: influence of head bone mineral density. J Bone Miner Res. 1997;12:652–655. doi: 10.1359/jbmr.1997.12.4.652. [DOI] [PubMed] [Google Scholar]
  • 27.Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United Stales measure by accelerometer. Med Sci Sports Exerc. 2008;40:181–8. doi: 10.1249/mss.0b013e31815a51b3. [DOI] [PubMed] [Google Scholar]
  • 28.Trost SG, Way R, Okely AD. Predictive validity of three ActiGraph energy expenditure equations for children. Med Sci Sports Exerc. 2006;38:380–387. doi: 10.1249/01.mss.0000183848.25845.e0. [DOI] [PubMed] [Google Scholar]
  • 29.United States Department of Health and Human Services (Internet) Physical Activity Guidelines for Americans. (cited 2008.) Available from http://www.hhs.gov/news/facts/physicalactivityguidelines.html.
  • 30.Vuori I, Heinonen A, Sievanen H, Kannus P, Pasanen M, Oja P. Effects of a unilateral strength training and detraining on bone mineral density and content in young women: A Study of mechanical loading and deloading on human bones. Calcif Tissue Int. 1994;55:59–67. doi: 10.1007/BF00310170. [DOI] [PubMed] [Google Scholar]
  • 31.Warden SJ, Durhs RK, Castillo AB, Nelson IR, Turner CH. Exercise when young provides lifelong benefits to bone structure and strength. J Bone Miner Res. 2007;22:251–259. doi: 10.1359/jbmr.061107. [DOI] [PubMed] [Google Scholar]
  • 32.Welk GJ, Corbin CB, Dale D. Measurement issues in the assessment of physical activity in children. Res Q Exerc Sport. 2000;71:59–73. doi: 10.1080/02701367.2000.11082788. [DOI] [PubMed] [Google Scholar]

RESOURCES