Table 1.

Characteristics of the eight observational studies on associations between sugar-sweetened beverage consumption and bone health in children^a,b

First author years (Ref) location	Study design	Sample size	Age or age range (Mean age ± SD)	Sex, % F	Sugar-sweetened beverages			Bone health			Main finding
					Method of assessment	Beverage category	Intake level	Method of assessment	Sites	Outcomes
Albala 2008 [29] c Chile	Randomized controlled trial	98	8–10 y	46.9	Modified FFQ	Sugar-sweetened beveragesd	Low: 742.8 ± 207.9e High: 802.1 ± 142.0, p = 0.10 g/d	DEXA	WB	Bone Mass	∙No difference in whole body bone mass between children fed different amounts of sugar-sweetened beverages (p = 0.56).
Fisher 2004 [30] c USA	Longitudinal study	182	9 y	100	24-h dietary recall	Sweetened beveragesf	Low: 358 High: 403, g/d	DEXA	WB	BMD	∙Girls who drank more sweetened beverages (p < 0.01) had a significantly lower whole body BMD (p < 0.001).
Libuda 2008 [31] Germany	Longitudinal study	228	6–18 y	49.6	3-day food records	Soft drinks	8y, Prepubescent Girls: 119.8 ± 129.2 Boys: 136.8 ± 137.3 13y, Pubescent Girls: 186.0 ± 196.5 Boys: 243.5 ± 200.4, g/d	pQCT	Forearm	BMC	∙Soft drinks consumption in children and adolescents was inversely associated with BMC at forearm (p = 0.036).
Ma 2004 [32] Australia	Case-control study	390	9–16 y	–	Questionnaire developed by author	Carbonated or cola drinks	Not reported	DEXA	WB LS FN	BMD	∙No significant correlation was shown between carbonated and/or cola drinks and bone measures, although all were inverse trends.
Manias 2006 [33] England	Case-control study	100	4–16 y	50	FFQ	Carbonated beverages	Low: 0.13 ± 0.17 High: 0.33 ± 0.57, p = 0.0182, ℓ/d	DEXA	LS UB LB	BMD BMC	∙Children who consumed more carbonated drinks (p = 0.0182) had a significantly lower BMD and BMC z-score at spine (BMD, p = 0.0003; BMC, p = 0.001), upper body (BMD, p = 0.015; BMC, p < 0.0001) and lower body (BMD, p = 0.015; BMC, p = 0.001).
McGartland 2003 [34] England	Cross-sectional study	1335	12–15 y	55.7	Dietary history	Carbonated soft drinksg	12y: Girls: 351 ± 332 Boys: 459 ± 394, p < 0.01 15y: Girls: 340 ± 380 Boys: 518 ± 452, p < 0.01, g/d	DEXA	DR HL	BMD	∙A significant inverse relationship between total intake of carbonated soft drinks and BMD was observed in girls at the forearm (p < 0.05) and heel (p < 0.05).
Nassar 2014 [35] Eqypt	Case-control study	100	Low: 10.3 ± 1.4y High: 10.6 ± 1.3y	44.1	Questionnaire developed by author	Sugar-sweetened beveragesd	Low: 1.08 ± 0.64 High: 3.16 ± 0.37, p < 0.001, number of intake /dayh	DEXA	LS	BMD	∙Children who consumed more than 12 oz had a significantly lower BMD (p < 0.001) than those that does not exceed 0–8 oz.
Whiting 2001 [36] Canada	Cross-sectional study	112	10–16 y	47.3	24-h recall	Carbonated and low nutrient-density beveragesi	Girls Carbonated, 96 ± 102 Low nutrient dense, 240 ± 177 Boys Carbonated, 246 ± 300 Low nutrient density, 429 ± 393, mL/d	DEXA	WB	BMC	∙Consumption of carbonated (p = 0.05) and low nutrient dense beverages (p = 0.03) was inversely related to BMC in adolescent girls but not in boys.

^aBMC bone mineral content, BMD bone mineral density, d day, DEXA Dual-energy X-ray absorptiometry, DR Distal radius, F Female, FFQ Food frequency questionnaire, FN Femoral neck, High high intake of SSBs, HL Heel, LB Lower body, Low low intake of SSBs, LS Lumbar spine, pQCT peripheral quantitative computed tomography, QUS Quantitative ultrasonography, Ref Reference, SD Standard deviation, SOS speed of sound, WB Whole body, UB Upper body, USA United States of America; y, year

^bQuality assessment was performed using the Cochrane criteria and Handel’s-developed scale and assessed by two authors (HA and YKP)

^cStandard errors (SE) presented in the original articles were converted to standard deviations (SD) for meta-analysis. This formula was used for conversion: SD=SEX√n

^d'Sugar-sweetened beverage’ included carbonated beverages and juice drinks-made by adding packaged sugary powders with fruit flavoring to water

^eValues are means ± SD (all such values)

^f ‘Sweetened Beverage’ included both energy-containing carbonated (soda) and noncarbonated beverages (fruit drinks, sport drinks, sweetened iced tea) that contained little if any fruit juice

^g'Carbonated soft drinks’ were defined as all nonalcoholic carbonated beverages that contained artificial sweeteners instead of added sugar

^h First, children with a daily intake of more than 12 oz of SSBs or less than 0–8 oz of SSBs were recruited. After that, the number of drinks they consumed per day was investigated

^i'Carbonated beverages’ includes cola, diet cola and other soft drinks; and ‘low nutrient dense beverages’ is the sum of carbonated and noncarbonated beverages. The latter included sugared drinks such as iced tea, Koolaid, coffee< 50% fruit juice, and fruit punches