Table 1

Skeletal loading exercise intervention studies in children and adolescents

Reference	Participants and design	Exercise intervention	Analytical approach	Results (% difference in gain between Ex and Con, after any statistical adjustments)
Con, Controls; Ex, Exercisers; TS, Tanner Stage; ROI, region of interest; ANOVA, analysis of variance; ANCOVA, analysis of covariance; TB, Total body; LS, Lumbar spine; PF, Proximal femur; FN, Femoral neck; GT, Greater trochanter; BMC, bone mineral content (by dual energy x ray absorptiometry (DEXA)); aBMD, areal bone mineral density; vBMD, volumetric bone mineral density (estimated from DEXA).
†Calculated eight month change from reported % change/month.
‡Bradney et al² and Witzke et al⁵⁸ were the only two studies that measured and reported change at the femoral midshaft (DEXA).
§Heinonen et al⁵ was the only study that measured and reported change at the tibial midshaft (pQCT).
Prepubertal
Bradney et al²	Boys. White.	Programme: One PE teacher supervised intervention programme	Con and Ex matched for age, standing	†TB: +1.2% aBMD.
	n=20 Con, n=20 Ex; mean age 10.4 (0.2); age	outside of school time. Included aerobics, soccer, volleyball, dance,	height, sitting height, weight, baseline	LS: +2.8% aBMD.
	range 8.4-11.8 years.	gymnastics, basketball, weight training.	aBMD. Unpaired t tests to compare bone	PF: Not measured.
	All TS 1 throughout.	Frequency & duration: 3 times/week; 30 minutes/session; 8 months	changes between groups.	FN: Not measured.
	Randomised by school: 1 Con + 1 Ex.	duration.		GT: Not measured.
		Progression: Not stated.		‡Femoral Mid-shaft: +5.6% BMC; +5.6% aBMD, vBMD NS, +6.4% cortical thickness.

Fuchs et al⁴	Girls & Boys. White.	Programme: Intervention took place outside of regular PE classes,	One factor ANCOVA; baseline bone,	TB: Not measured.
	n=44 Con; n=45 Ex; mean age 7.5 (0.2); age	supervised by research team. Each session: 100 two footed drop	change in height, change in weight, and	LS: +3.1% BMC; +2.0% aBMD.
	range 5.9-9.8 years. All TS 1 throughout.	landings from 61 cm height on to a wooden floor. Mean (SD) ground	age as covariates.	PF: Not reported.
	Children were randomised to Con or Ex within one	reaction force for drop jumps from 61 cm height was 8.8 (0.9) times		FN: +4.5% BMC; aBMD not significant.
	elementary school.	body weight.		GT: Not reported.
		Frequency & duration: 3 times/week; 10 minutes jumping/session; 7 months.
		Progression: Week 1-4: Progressed from 50 jumps per session (no box), to 80 jumps per session (from box). Week 5-end : 100 jumps from 61 cm box.

McKay et al³	Girls & Boys. Asian & White.	Programme: School based; teachers chose activities from a variety of	2 (Ex, Con) × 2 (Asian, White) × 2	TB: Not significant.
	n=81 Con; n=63 Ex; mean age 8.9 (0.7) years; age	games, circuits, dances which incorporated jumping. 10 tuck jumps	(male, female) ANOVA to examine bone	LS: Not significant.
	range 6.9-10.2 years. Boys were TS 1 throughout;	performed before PE class, and once in classroom each week.	changes between groups. Hierarchical	PF: Not significant.
	89% of girls were TS1 and 11% were TS 2 at	Frequency & duration: 3 times/week; 10-30 minutes/session; 8	regression to examine effect of	FN: Not significant.
	baseline; ~30% of girls were TS 2 by follow up.	months.	intervention when baseline bone, height	GT: BMC not significant; +1.4% aBMD.
	Randomised by school: 5 Con + 5 Ex.	Progression: As per fitness level of class; more challenging activities	change, lean mass change, general
		added as options after 3 months.	physical activity, calcium, sex, and ethnicity controlled.

Early pubertal
Morris et al¹	Girls. Ethnicity not stated, schools matched for	Programme: One PE teacher supervised intervention programme	Independent t tests to compare bone	TB: +5.5% BMC; +2.3% aBMD.
	ethnicity.	outside of school time. Included aerobics, soccer, football, step	change between groups. ANCOVA (with	LS: BMC not significant; +3.6% aBMD;
	n=33 Con; n=38 Ex girls; mean age 9.5 (0.9)	aerobics, dance, skipping, ball games, weight training.	change in height and total body mass as	+2.9% vBMD.
	years; age range 9-10 years. All premenarcheal	Frequency & duration: 3 times/week; 30 minutes/session; 10 months.	covariates) to compare adjusted change	PF: BMC not significant; +3.2% aBMD.
	throughout; TS 1-3.	Progression: In 10 week weight training session only.	in bone between groups.	FN: +4.5% BMC; aBMD not significant;
	Not randomised; schools self selected: 1 Con + 1			vBMD not significant.
	Ex.			GT: Not reported.

Adolescents
Blimkie et al⁵⁷	Girls. Ethnicity not stated.	Programme: Resistance training using hydraulic machines (13	Con and Ex matched for age, body	TB: Not significant.
	n=16 Con, n=16 Ex girls.	exercises, 4 sets with 10-12 reps of each). Sessions supervised by	mass, level of habitual physical activity.	LS: Not significant.
	Mean age 16.2 (0.2) years; age range 14-18	researchers.	Two way repeated measures ANOVA	PF: Not measured.
	years. All postmenarcheal (TS 4-5) at baseline.	Frequency & duration: 3 times/week; session duration not stated; 6.5	used to compare change in bone	FN: Not measured.
	Girls were randomised to Con or Ex within 1	months.	between groups.	GT: Not measured.
	highschool.	Progression: Resistance increased every 6 weeks.

Witzke et al⁵⁸	Girls. All white.	Programme: First 3 months: resistance training + plyometrics. Next 6	Repeated measures ANOVA.	TB: Not significant.
	n=29 Con; n=27 Ex; mean age 14.6 (0.5) years;	months: plyometrics, including jumps, depth jumps, bounding and		LS: Not significant.
	age range13-15 years. All postmenarcheal at	hopping on soft surfaces.		PF: Not significant.
	baseline. Con and Ex matched for age and months	Frequency & duration: 3 times/week; 30-45 minutes/session; 9		FN: Not significant.
	postmenarche.	months.		GT: Not significant.
	Not randomised; exercisers participated for PE credit	Progression: Weight training progressed from months 1-3:		‡Femoral Mid-shaft: Not significant.
	at 2 highschools.	repetitions, sets, and weight gradually increased. Plyometric training
		progressed in jump difficulty and number of reps.
Two maturity groups
Heinonen et al⁵	Girls. All white.	Programme: Jump training sessions incorporating single and two foot	Individual BMC values normalised by the	Premenarcheal:
	n=58 (33 Con + 25 Ex) Premenarcheal (TS 1-3,	jumps from floor, and on and off a 30 cm box.	length of the ROI. ANCOVAs performed	TB: Not measured.
	mean age 11.0 (0.9) years (Con), 11.7 (1.3) years	Frequency & duration: 2 times/week; 20 minutes jump	within each maturity group, (baseline	LS: +3.3% BMC.
	(Ex)).	training/session; 9 months.	bone values and age as covariates).	PF: Not significant.
	n=68 (29 Con + 39 Ex) Postmenarcheal (TS 2-5;	Progression: Progressed gradually from two foot floor jumps (100		FN: +4.0% BMC.
	mean age 13.7 (0.9) years (Con & Ex)).	jumps) to combination one and two foot jumps from box (200 jumps).		GT: Not significant.
	Not randomised; schools self-selected to 2 Ex + 3			§Tibial Mid-shaft: Not significant.
	Con.			No significant differences between postmenarcheal Ex and Con.

MacKelvie et al⁶	Girls. Mixed ethnicities (primarily white + Asian).	Programme: School based, cicuit training programme, integrated into	ANCOVAs within maturity categories to	No significant differences between
	n=70 (26 Con + 44 Ex) Prepubertal (TS 1, mean	physical education classes, supervised by classroom teachers. All	compare bone change between groups	prepubertal Ex and Con.
	age 10.1 (0.5) years).	station activities incorporated jumping activities: i.e., drop jumps, star	(controlling for age, maturity, baseline	Early pubertal:
	n=107 (64 Con + 43 Ex) Early pubertal (TS 2 + 3,	jumps, lunge jumps, hopping, plyometric jumps, jumping over	bone values, change in height and/or	TB: Not significant.
	mean age 10.5 (0.6) years).	obstacles. Average ground reaction force for various jumps ranged	bone area).	LS: +1.8% BMC; +1.7% aBMD.
	Randomised by school: 7 Ex + 7 Con.	between 3.5-5 times body weight.		PF: Not significant.
		Frequency & duration: 3 times/week; 12 minutes/session; 7 months.		FN: +1.9% BMC; +1.6% aBMD; +3.1%
		Progression: Levels 1-3 (~2 1/2 months at each level). Number of		vBMD.
		jumps increased within level (50-100); and jump difficulty/height increased between levels.		GT: Not significant.