Table 14.

Physical activity and exercise on bone structure in children and adolescents

Reference	Study description	Population description	Number of subjects	Intervention	End points	Results
RCTs
Heinonen et al. 2000 [244]	9-month step aerobics/ jumping intervention Randomized by school	Sex: female Age: 10–15 years Race: not specified Location: Tampere, Finland	126	Frequency: 2×/week Intensity: progressive Time: 50 min (10-min warm-up, 15-min nonimpact aerobic exercises, 20-min high-impact/jump training, 5-min cool down) Type: both-leg jumps at floor level, both-leg box jumps, one-leg box jumps Control: normal PA	pQCT Significant None NS Tibia CoD, CoA, BSI	At the tibial midshaft, the intergroup differences (CoD, CoA, and BSI) were not significant
Petit et al. 2002 [73]	7-month jumping intervention Randomized by school	Sex: female Age: 9–12 years Race: multiethnic population of city (~34 % Hong Kong Chinese and 57 % white) reflected in cohort Location: Richmond, British Columbia	177	Frequency: 3×/week (2× PE, 1 classroom) Intensity: progressed over school year by increasing number of jumps per station (from 10 to 20) and height (from 10 to 50 cm) Time: 10–12 min (5 circuits, ~1.5–2 min each) Type: jumping exercises—jumping jacks, lung jumps, hopping, jumping over various obstacles, drop jumps from platform Control: stretching	pQCT	Change in early pubertal girls, mean (95 % CI)
					Significant	EX		CON
					NN CSA	0.060 (0.043, 0.077)*		0.035 (0.021, 0.049)
					NN Z	0.145 (0.118, 0.0172)*		0.106 (0.084, 0.129)
					NN CT	0.012 (0.008, 0.016)*		0.007 (0.004, 0.010)
					IT ED	0.092 (0.064, 0.120)*		0.138 (0.114, 0.161)
					NS	There was no difference in change for bone structure in the PRE girls. The more mature girls (EARLY) in the EX group showed significantly greater gains in CSA and reduced endosteal expansion. Changes in SPW did not differ. Structural changes improved Z at the NN (4.0 %), but not at the IT region. There were no differences at the primarily cortical FS.
					NN length, SPW, ED
					IT CSA, SPW, Z
					FS CSA, SPW, Z, ED, CT
Specker and Binkley 2003 [81]	1-year randomized, placebo-controlled, partially blinded trial of PA and calcium supplementation	Sex: male and female Age: 3–5 years Race: 94 % white, 6 % other Location: South Dakota	178	Frequency: 5 days/week Intensity: not specified Time: 30 min/day (5-min warm-up, 20-min activity, 5-min cool down) Type: hopping, jumping, skipping activities (17 different weekly programs) Control: 30 min/day of activities to keep them sitting quietly	pQCT	Exercise group had greater PCirc and ECirc than control group (P < 0.05).
					Significant
					PCirc
					ECirc
					NS
					CA, CT
MacKelvie et al. 2004 [295]	20 month jumping intervention Randomized by school	Sex: male Age: 8.8–12.1 years Race: multiethnic population of city (~34 % Hong Kong Chinese and 57 % North American/Western European Caucasian, 5 % Southeast Asian, and 4 % other or mixed ethnicity) reflected in cohort Location: Richmond, British Columbia	64	Frequency: 3×/week (2× PE, 1 classroom) Intensity: progressed over school year by increasing number of jumps per station (from 10 to 20) and height (from 10 to 50 cm Time: 10–12 min (5 circuits, ~1.5–2 min each) Type: jumping exercises- jumping jacks, lung jumps, hopping, jumping over various obstacles, drop jumps from platform Control: stretching	HSA	Change, mean (95 % CI)
					Significant	EX		CON
					NN CSMI	0.23 (0.20, 0.27)*		0.17 (0.14, 0.20)
					NN Z	0.13 (0.11, 0.15)*		0.09 (0.072, 0.11)
					NS	At the NN region, EX boys had significantly greater changes in Z (7.5 %). Changes at the IT and FS regions were not significantly different between groups.
					NN length, CSA, PW, EW, CT
					IT CSMI, CSA, PW, Z, ED, CT
					FS CSA, PW, CSMI, Z, EW, CT
McKay et al. 2005 [297]	8-month jumping intervention Control group used from previous study	Sex: male and female Age: 8.9 to 11.0 years Race: 38 % Caucasian, 48 % Asian, 15 % other (including mixed ethnic, black, and South Asian) Location: Richmond, British Columbia	124	Frequency: 3×/day Intensity: 5× body wt, maximum rate of force was > 400 body wt/s (independent sample of 70 boys and girls) Time: 3 min/day Type: 10 counter movement jumps (two-foot take off, clutch knees, two-foot landing) Control: not specified, from previous study	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					NN and FS BMD, CSA, SW, Z, ED, CT
Macdonald et al. 2007 [25]	16-month PA and jumping intervention Randomized by school	Sex: male and female Age: 10.2 ± 0.6 years Race: 53 % Asian (Hong Kong or China, India Philippines, Vietnam, Korea, or Taiwan), 35 % Caucasian (North America or Europe), 12 % mixed or other ethnic origins Location: Richmond and Vancouver, British Columbia	410	Frequency: Classroom Action: 5 days/week; Bounce at the Bell: 5 days/week PA + 3×/day, 4 days/week jumping Intensity: not specified Time: Classroom Action: 15 min of PA; Bounce at the Bell: 15 min of PA + short bouts of high-impact jumping Type: Classroom Action: skipping, dancing, playground circuits and simple resistance exercises with bands; Bounce at the Bell: Classroom Action + short bouts high-impact jumping Control: regular PE	pQCT	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					Distal and midshaft BSI, ToA, ToD
Linden et al. 2007 [298]	1 year expanded PE intervention Randomized by school	Sex: Male Age: 6.7–9.0 years Race: Caucasian (except 1 boy adopted from Colombia) Location: Malmö, Sweden	138	Frequency: 5 days/week Intensity: varied Time: 40 min/day Type: normal PE curriculum (running, jumping, climbing ropes, variety of ball games). Activities varied to reduce boredom. No specific osteogenic training program was added. Control: normal PE curriculum, duration within normal limits (1–2 sessions/week, in total 60 min/week)	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					FN CSMI, Z, CSA
Macdonald et al. 2008 [299]	8-month PA and jumping intervention Randomized	Sex: male and female Age: 9–11 years Race: 53 % Asian (both parents or all four grandparents born in Hong Kong or China, India, Philippines, Vietnam, Korea, or Taiwan. 35 % Caucasian (parents born in North America or Europe). 12 % children mixed ethnicity or other ethnic origins Location: Vancouver and Richmond, British Columbia	410	Frequency: Classroom Action: 5 days/week; Bounce at the Bell: 5 days/week PA + 3×/day, 4 days/week jumping Intensity: not specified Time: Classroom Action: 15 min PA; Bounce at the Bell: 15 min PA + short bouts high-impact jumping Type: Classroom Action: skipping, dancing, playground circuits, simple resistance exercises with bands; Bounce at the Bell: Classroom Action + short bouts high-impact jumping. Phase I included 5 two-foot landing jumps (or 10 one-foot landing jumps) at each session. Phase II jumps were increased each month of the school year until a maximum of 36 jumps/day was reached. Control: Usual PE	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					NN Z, CSA, SPW
Alwis et al. 2008 [300]	1-year expanded PE intervention Randomized by school	Sex: female Age: 6.5–8.9 years Race: Caucasian Location: Malmö, Sweden	103	Frequency: 5 days/week Intensity: varied Time: 40 min/day Type: normal PE curriculum (running, jumping, climbing ropes, variety of ball games). Activities varied to reduce boredom. No specific osteogenic training program was added. Control: normal PE curriculum, duration within normal limits (1–2 sessions/week, in total 60 min/week)	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					FN length, PW, CSA, Z, CSMI, EW, CT
Alwis et al. 2008 [300]	2-year expanded PE intervention Randomized by school	Sex: male Age: 6.7–9.0 years Race: Caucasian Location: Malmö, Sweden	137	Frequency: 5 days/week Intensity: varied Time: 40 min/day Type: normal PE curriculum (running, jumping, climbing ropes, variety of ball games). Activities varied to reduce boredom. No specific osteogenic training program was added. Control: normal PE curriculum, duration within normal limits (1–2 sessions/week, in total 60 min/week)	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					FN CSA, Z, CSMI
Alwis et al. 2008 [300]	2-year expanded PE intervention Randomized by school	Sex: female Age: 6.8–8.9 years Race: Caucasian Location: Malmö, Sweden	291	Frequency: 2×/week (PE); 3×/day (activity breaks); 1×/day (home activity) Intensity: not specified Time: 45 min/session (PE); 2–5 min/session (activity breaks); 10 min/day (home activity) Type: 2 PE lessons (plus 3 regular PE classes) taught mostly outdoors by PE teachers. All five sessions included jumping activities such as hopping, jumping up and down stairs, rope skipping, etc. During academic lessons, 3–5 activity breaks comprised motor skill tasks such as jumping around on one leg, balancing on one leg, power games, or coordinative tasks were introduced every day. Daily home activity included aerobic, strength, or motor skill tasks such as tooth brushing while standing on one leg, jumping up and down the stairs, and rope jumping. Control: participated in regular PE classes 3×/week	HSA	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					FN CSA, PW, Z, CSMI
Weeks et al. 2008 [296]	8-month jumping intervention Randomized	Sex: male and female Age: boys 13.8 years (0.4); girls 13.7 years (0.5) Race: not specified Location: Gold Coast, Australia	81	Frequency: 2×/week Intensity: 1–3 Hz at a height of 0.2–0.4 m Time: 10 min (~300 jumps) Type: varied, but included- jumps, hops, tuck jumps, jump squats, stride jumps, star jumps, lunges, side lunges, and skipping. Occasionally supplemented with upper body strengthening activities, including push-ups and exercises with resistive latex bands Control: usual PE warm-up and stretching	DXA	Percent change, mean
					Significant	EX		CON
					LS IBS	17.9*		14.4
					NS	LS IBS improved more for EX participants than for CON. No other differences in 8-month change in parameters of bone strength reached significance.
					FN, LS BMAD
					FN CSMI
Greene et al. 2009 [301]	28-week drop-landing intervention Randomized	Sex: female Age: 6–10 years Race: not listed Location: Sydney, Australia	39	Frequency: 3×/week Intensity: 14-cm height drop for low drop, 28-cm height drop for high drop (GRF 2–4× body wt); 10 sets of 5 steps per set; 50 steps per session Time: 15 min/day Type: single leg drop-landing exercises with nondominant leg serving as the training leg (dominant leg served as within-participant untrained control) Control: normal activity	pQCT	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					Cort, Medullary, and Total BA at Proximal, Mid, Distal
Macdonald et al. 2009 [250]	16-month PA and jumping intervention Randomized	Sex: male Age: 10.2 ± 0.6 years Race: 52 % Asian (Hong Kong or China, India, Philippines, Vietnam, Korea, or Taiwan), 35 % Caucasian (North America or Europe), 12 % mixed or other ethnic origins Location: Richmond and Vancouver, British Columbia	205	Frequency: Classroom Action: 5 days/week; Bounce at the Bell: 5 days/week PA + 3×/day, 4 days/week jumping Intensity: not specified Time: Classroom Action: 15 min PA; Bounce at the Bell: 15 min PA + short bouts high-impact jumping Type: Classroom Action: skipping, dancing, playground circuits, simple resistance exercises with bands; Bounce at the Bell: Classroom Action + short bouts high-impact jumping. Phase I included 5 two-foot landing jumps (or 10 one-foot landing jumps) at each session. Phase II jumps were increased each month of the school year until a maximum of 36 jumps/day was reached. Control: Usual PE	pQCT	Difference, mean (95 % CI)
					Significant
					I max	338.6 (16.4, 660.9)*
					NS	The EX boys had a 3 % greater gain in I max than CON boys. No other bone structural parameters differed between groups.
					I min, I max/I min, CoA- Ant, CoA- Med, CoA- Post, CoA- Lat, CTh- Ant, CTh- Med, CTh- Post, CTh- Lat
Lofgren et al. 2012 [28]	4-year expanded PE intervention Randomized by school	Sex: male and female Age: 6.5–8.7 years Race: Caucasian Location: Malmö, Sweden	221	Frequency: 5 days/week Intensity: varied Time: 40 min/day Type: normal PE curriculum (running, jumping, climbing ropes, variety of ball games). Activities varied to reduce boredom. No specific osteogenic training program was added Control: normal PE curriculum, duration within normal limits (1–2 sessions/week, in total 60 min/week)	HSA	Change, mean (95 % CI)
						Male	Female
						EX	CON
					FN CSA		0.11 (0.10, 0.13)*	0.09 (0.07, 0.11)
					FN Z		0.06 (0.05, 0.06)*	0.05 (0.04, 0.06)
					FN CSMI		0.10 (0.08, 0.11)*	0.08 (0.07, 0.09)
						The EX girls had significantly greater gain in all HSA outcomes than CON girls. No differences were seen for boys.
Anliker et al. 2012 [302]	9-month jumping intervention Randomized by school	Sex: male and female Age: 8–12 years Race: not listed Location: Lucerne, Switzerland	45	Frequency: 2×/week Intensity: intensity increased weekly. Number of jumps increased over time from ~60 to ~150 per session Time: 10 min Type: Jumping and sprinting exercises, including two- and one-legged hopping, drop jumps, side to side jumps, jumping jacks, jumps and landings from a podium, jumps over barriers and short multidirectional sprints Control: not specified	pQCT	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					Tibia 4 % vBMC, vBMDtr, vBMDtot, BAtb, BAtot
Detter et al. 2014 [303]	6-year expanded PE intervention Randomized by school	Sex: male and female Age: 6–8 years Race: Caucasian Location: Malmö, Sweden	133	Frequency: 5 days/week Intensity: varied Time: 40 min/day Type: normal PE curriculum (running, jumping, climbing ropes, variety of ball games). Activities varied to reduce boredom. No specific osteogenic training program was added Control: normal PE curriculum, duration within normal limits (1–2 sessions/week, in total 60 min/week)	Cross-sectional pQCT	Change in bone structural parameters did not differ between groups.
					Significant
					None
					NS
					Tibia 4 %, Radius 4 %
					TR vBMD
					Tibia 38 %, Radius 66 %
					Cort vBMD
					Cort BMC
					Cort BA
					Cort Th
					CSA
					Polar SSI
Observational studies
Reference	Study description	Population description	N	Exposure	End points	Results
Forwood et al. 2006 [363]	7-year prospective follow-up	Sex: 109 males, 121 females Baseline age: 8–15 years Race: white Location: Saskatoon, Saskatchewan, Canada	230	Method: self-report questionnaire (3×/year for 3 years, 2×/year after) Variable: general level of PA for year (1–7)	Multilevel models for FN Z, CSA, and SPW (HSA)	MVPA positively associated with FN Z and FN CSA males and females. A male with PA score of 1 had 0.0972 cm2 less CSA than a male with PA score of 5. Females with PA score of 1 had 0.0588 cm2 less CSA than females with score of 5. Males with PA score of 1 had 0.0615 cm2 less Z than males with PA score of 5. Females with PA score of 1 had 0.0355 cm2 less Z than females with score of 5.
Janz et al. 2007 [364]	6-year prospective follow-up	Sex: 212 males, 233 females Baseline age: 5 years Race: primarily white Location: Iowa, USA	445	Method: ActiGraph accelerometer Variable: MVPA min/day (>3000 ct/min)	Multilevel models for FN Z and CSA (HSA)	MVPA positively associated with Z and CSA males and females. When adjusted, LM only significant males
						40-min MVPA predicted 3–5 % greater CSA and Z compared to 10 min/day MVPA
						When CSA and Z adjusted for LM, associations attenuated in males but remain significant. Associations become insignificant in females.
Farr et al. 2013 [360]	2-year prospective follow-up	Sex: 248 females Baseline age: 9–12 years Race: 90 % white, 6 % Asian, 2 % black or African American, 0.5 % Native American or Alaska Native, 1 % Native Hawaiian or Pacific Islander, 0.5 % other Location: Tucson, Arizona, USA	248	Method: self-report questionnaire Variable: past-year PA score	Follow-up tibia and distal femur BSI	PA associated with increases in distal femur BSI
Gruodyte-Raciene et al. 2013 [365]	4-year prospective follow-up	Sex: 81 males, 84 females Baseline age: 4–10 years Race: 96 % white, 2 % Asian, 2 % other Locations: Saskatoon, Saskatchewan, Canada	165 92 gymnasts, 73 nongymnasts	Method: parental-report questionnaire Variable: PA score 1–2 h/week gymnastics exposure	Yearly trajectories of PF CSA and Z (HSA)	When compared to nongymnasts, gymnasts 6 % greater NN CSA, 7 % greater NN Z, 5 % greater IT CSA, 6 % greater IT Z, and 3 % S CSA
Janz et al. 2014 [248]	12-year prospective follow-up	Sex: 160 males, 189 females Baseline age: 5 years Race: primarily white Location: Iowa, USA	349	Method: ActiGraph accelerometer Variable: MVPA min/day (≥2296 ct/min)	Follow-up (age 17) FN Z and CSA Tibia BSI Tibia polar moment of inertia	Participants in high trajectory for MVPA had greater geometric outcome measures (Z: females 21.0 %, males 9.4 %; CSA: females 11.8 %, males 10.0 %; BSI: females 19.0 %, males 12.8 %; polar moment of inertia: females 38 %, males 8.4 %) at age 17 compared to other trajectories.
Jackowski et al. 2014 [251]	~15-year prospective follow-up	Sex: 55 males, 49 females Baseline age: 8–15 years Race: primarily white Location: Saskatoon, Saskatchewan, Canada	104	Method: self-report questionnaires Variable: average activity score	Follow-up (age 23–30) PF CSA and Z (HSA)	Around time of PHV, active adolescents had 8–12 % greater CSA and 9–12 % greater Z than inactive peers at PF. When adult CSA and Z were adjusted for height, weight, adolescent bone geometry, and sex, those active during adolescence maintained 5–7 % benefit in CSA and 6–8 % benefit in Z in adulthood compared to inactive. In young adulthood, males and females who were physically active as adolescents had greater bone geometric measures.
Duckham et al. 2014 [252]	~18-year prospective follow-up	Sex: 49 males, 73 females Baseline age: 8–15 years Race: primarily white Location: Saskatoon, Saskatchewan, Canada	122	Method: self-report questionnaire Variable: general PA score	Follow-up (age 24–34) tibia and radius BA, BSI, SSIp, CoC, CoA, CoD, ToA, ToD, TrC, TrD, and BSIc (pQCT)	In young adulthood, males who were physically active as adolescents had 13 % greater SSIp and 10 % greater ToA at tibia. Females had 10 % larger CoA, 12 % larger CoC, and 3 % larger TrC at tibia.
Cardadeiro et al. 2014 [362]	1-year prospective follow-up	Sex: 81 males, 96 females Baseline age: 10–12 years Race: primarily white Location: Lisbon area, Portugal	177	Method: bone-specific administered questionnaire ActiGraph accelerometer Variables: PA score Sedentary min/day (≤100 ct/min) Light min/day (101–2295 ct/min) Moderate min/day (2296–4011 ct/min) Vigorous min/day (>4012 ct/min)	BMD distribution via three aBMD ratios: FN/PF, IM/SL, and TR/PF Geometric measures of the pelvis—IAD and PF ALA	Questionnaire significant in FN/PF ratio model in males

95 % CI 95 % confidence interval, aBMD areal bone mineral density (g/cm²), ALA abductor lever arm, Ant anterior, BA bone area (cm²), BMAD bone mineral apparent density (g/cm²), BMC bone mineral content (g), BMD bone mineral density (g/cm²), BSI bone strength index (mg²/cm⁴), BSI _c bone strength in compression, CA cortical area (mm²), CoA cortical bone area (mm²), CoC cortical content (mg/mm), CoD cortical density (mg/cm³), CON control group, Cort cortical, CSA cross-sectional area (cm²), CSMI cross-sectional moment of inertia, CT cortical thickness (mm), ct cortical, ECirc (mm) endosteal circumference, ED endosteal diameter (cm), endo endosteal, EW endosteal width, EX exercise group, FN femoral neck, FS femoral shaft, HSA hip structure analysis, IAD interacetabular distance, IBS index of bone structural strength, IM inferomedial, I _max maximum moment of area, I _min minimum moment of area, IT intertrochanter, Lat lateral, LM lean mass, LS lumbar spine, Med medial, MVPA moderate-through-vigorous-intensity physical activity, NN narrow neck, NS not significant, PA physical activity, PCirc periosteal circumference (mm), PE physical education, peri periosteal, PF proximal femur, PHV peak height velocity, Post posterior, pQCT peripheral quantitative computed tomography, PW periosteal width, RCT randomized controlled trial, SL superolateral, SPW shaft periosteal width (cm), SSI strength strain index, SSIp density-weighted polar section modulus (mm³), SW subperiosteal width (cm), tb trabecular, Th thickness, ToA total area (mm²), ToD total BMD (mg/cm³), tot total, TR trabecular, TrC trabecular content, vBMC volumetric BMC (g/cm), vBMD volumetric bone mineral density (g/cm³), Z section modulus (cm³)

*P < 0.05; **P < 0.01; ***P < 0.001