Summary of findings for the main comparison. Interventions to reduce the consumption of SSB compared to no or alternative intervention: impact on SSB intake.
| Interventions to reduce the consumption of SSB compared to no or alternative intervention: impact on SSB intake (additional outcomes reported in Summary of Findings tables 2‐9)a | ||
| Intervention type |
Outcomes (follow‐up) No. of clusters or participants No. of studies Certainty of Evidenceb |
Impact on direct and indirect measures of SSB intake |
| A Labelling interventions | ||
| A.1 Traffic‐light labelling |
SSB sales (12 months)
42 points‐of‐sale in 2 hospitals in 2 cities 2 ITS studies ⊕⊕⊕⊝ MODERATEc |
Boelsen‐Robinson 2017 (units of red‐labelled beverages sold):−56% (95% CI −67 to −45) Hartigan 2017 (share of red‐labelled beverages among all beverages sold): −25 percentage points (P < 0.001) |
| A.2 Nutritional rating score shelf‐labels |
SSB sales (7 to 11 months) 442 stores from 4 chains in 2 countries 1 ITS and 1 CBA study ⊕⊕⊝⊝ LOW |
Cawley 2015 (units of SSB sold):−27.4% (no P value or CI reported) Hobin 2017 (share of beverages with zero stars (mainly SSB), coefficient estimate): −0.026, P < 0.001 |
| A.3 Menu‐board calorie labelling |
Beverage calories per transaction (4 to 12 months)
353 stores from 4 chains in 6 cities 1 controlled ITS, 2 CBA studies ⊕⊝⊝⊝ VERY LOWd,e |
Bollinger 2011 (beverage calories per transaction): −0.3% (P < 0.01) Elbel 2013 (beverage calories per transaction):No effects (data not shown) Finkelstein 2011 (beverage calories per transaction): +1.7 kcal (95% CI −1.5 to 4.9) |
| A.4 Emoticon labelling |
Sugar‐sweetened milk selection (4 months)
186 students in 2 school cafeterias 1 ITS study ⊕⊕⊝⊝ LOW |
Siegel 2016a (share of students selecting chocolate milk): −16 percentage points (95% CI −27 to −4) |
| B Nutrition standards in public institutions | ||
| B.1 Reduced availability of SSB in schools |
SSB intake (6 to 24 months)
18,238 students in 240 schools 5 CBA studies ⊕⊕⊝⊝ LOW |
Cradock 2011 (total SSB intake): −99 ml/day (95% CI −173 to −26) Whatley Blum 2008 (total SSB intake): −14 ml/day (95% CI −69 to 41) Bauhoff 2014 cohort (share of students consuming any SSB): −4 percentage points (95% CI −10 to 2) Bauhoff 2014 crosssectional (share of students consuming any SSB): −10 percentage points (95% CI −17 to −3) Schwartz 2009 (total intake of beverages excluded by nutrition standards (mainly SSB) in schools):Decrease (ß = −0.23, P < 0.05) |
| B.2 Improved access to drinking water in schools |
SSB intake (3 to 13 months)
11,253 students in 62 schools 1 CBA study, 3 NRCTsf ⊕⊝⊝⊝ VERY LOWd,f,g |
Elbel 2015a (total SSB intake): No statistically significant effect (data not shown) Muckelbauer 2009 (total SSB intake, participants with foreign‐born parents or grandparents): −20 ml/day (95% CI −60 to 20) Muckelbauer 2009 (total SSB intake, participants without foreign‐born parents or grandparents):±0 ml/day (95% CI −60 to 60) Van de Gaar 2014 (total SSB intake): −190 ml/day (95% CI −280 to −100) Visscher 2010 (SSB sales at school):+0.7 ml/day/student (no statistical analyses reported by study authors) |
|
Sugar‐sweetened milk intake (5 years) 1,065,562 students in 1227 schools 1 CBA study ⊕⊝⊝⊝ VERY LOWd |
Schwartz 2016 (sugar‐sweetened milk intake): −3 ml/day (95% CI −5 to −1) | |
| B.3 Small prizes for the selection of healthier beverages in school cafeterias |
Sugar‐sweetened milk selection and purchases (3 to 20 months)
4213 students in 12 schools
(1 RCT and 2 ITS studies)
⊕⊕⊕⊝ MODERATEd (Hendy 2011) ⊕⊕⊝⊝ LOW (Emerson 2017; Siegel 2016b) |
Hendy 2011 (number of meals with unhealthy beverages selected) −3.0 meals/week (P = 0.000) Siegel 2016b (chocolate milk purchases): −0.12 servings/day (P < 0.001) Emerson 2017 (chocolate milk purchases): −0.12 servings/day (P < 0.001) |
| B.4 Improved placement of healthier beverages in school cafeterias | Sugar‐sweetened milk intake (3 to 4 months) 2638 students in 14 schools 1 cluster‐RCT ⊕⊕⊝⊝ LOWd,g | Cohen 2015 (selection and consumption of sugar‐sweetened milk): No statistically significant effect (data not shown) |
| B.5 Fruit provision in schools | SSB intake (9 months to 6 years) 3494 students in 47 schools 1 cluster‐RCT and 1 CBA study ⊕⊝⊝⊝ VERY LOWh,i,j | Da Costa 2014 (SSB intake): Decrease (P = 0.003, data not intelligible to review authors) Øverby 2012 (SSB intake measured with an unhealthy snack frequency score): −1.4 in the free‐fruit provision group, −1.1 in the fruit subscription group, −0.7 in the control group (P = 0.002 for time*group interaction) |
| C Economic tools | ||
| C.1 Price increases on SSB | SSB sales (4 to 12 months) 1 store, 7 leisure centres and 37 restaurants in 29 cities 3 ITS studies ⊕⊕⊕⊝ MODERATEk | Cornelsen 2017 (SSB items sold per customer): −9% (95% CI −15 to −3) Blake 2018 (volume of red‐labelled beverages sold): −28% (95% −32 to −23) Breeze 2018 (volume of SSB sold per attendance): −27% (95% CI −59 to −3) |
| C.2 Financial incentives to purchase low‐calorie beverages implemented through supermarket loyalty cards | SSB intake and sales (3 to 6 months) 1750 customers of 3 supermarket chains 3 RCTs ⊕⊕⊝⊝ LOWd,g |
Ball 2015 (self‐reported SSB intake): +10 ml/day (95% CI 0 to 20)
Ball 2015 (SSB purchases): +55 ml/day (95% CI −7 to 117)
Ni Mhurchu 2010 (energy density of beverages purchased): −0.1 MJ/kg (95% CI −0.4 to 0.2) Franckle 2018 (number of red‐labelled beverages purchased per month): −0.14 beverage items/month (95% CI −0.8 to 0.6) |
| C.3 Price discounts on low‐calorie beverages in community stores | SSB intake (6 months) 8515 inhabitants of 20 remote indigenous communities 1 cluster‐RCT ⊕⊕⊝⊝ LOWd,g | Brimblecombe 2017 (SSB sales): +6% (95% CI −3 to 15) |
| Taxation of SSB | Not included in this review (for a forthcoming Cochrane Review on taxation of SSB see Heise 2016, for existing systematic reviews on taxation of SSB see Backholer 2016, Cabrera Escobar 2013 and Nakhimovsky 2016) | |
| D Advertisement regulation: No studies found | ||
| E Whole food supply interventions | ||
| E.1 Voluntary food and beverage industry initiatives to improve the nutritional quality of the whole food supply | SSB sales and purchases ( 3 to 4 years) 61,126 households and 17 companies 2 controlled ITS and 1 CBA study ⊕⊝⊝⊝ VERY LOWd | Ng 2014a (energy from beverages sold):Decrease: −14 kcal/capita/day in the intervention group compared to −3 kcal/capita/day for national‐brand companies not participating in the pledge, no P value or CI reported) Ng 2014a (SSB sales by companies participating in the Healthy Weight Commitment Foundation Pledge): −7 kcal/per capita/day (no P Value or CIs reported, data for the CG not reported) Ng 2014b (SSB purchases): Increase relative to expected trends (P < 0.001, data shown graphically only) Taillie 2015 (percentage volume of SSB purchased):Decrease (P < 0.01, results shown graphically only) |
| F Retail and food service interventions | ||
| F.1 Healthier default beverages in children's menus in restaurants | SSB sales (6 years) 145 restaurants in 1 theme park 1 ITS study ⊕⊕⊝⊝ LOWc,d | Peters 2016a (share of children's menus served with SSB):−68 percentage points (no P value or CI reported for the pre‐post‐comparison) |
| F.2 In‐store promotion of low‐calorie beverages in supermarkets | SSB sales (6 months) 8 supermarkets from 2 chains 1 cluster‐RCT ⊕⊕⊕⊝ MODERATEg | Foster 2014 (in‐aisle SSB sales per supermarket): −11 l/day (95% CI −63 to 40) Foster 2014 (check‐out cooler SSB sales per supermarket): −2 units/day (95% CI −5 to 1) |
| F.3 Healthier vending machines in workplaces and schools | SSB intake (18 to 24 months) 6 schools and 4 worksites 2 NRCTsf ⊕⊝⊝⊝ VERY LOWd,e,j | Ermetici 2016 (SSB intake, normal‐weight participants): −1.1 times/week (95% CI −1.5 to −0.7) Ermetici 2016 (SSB intake, participants with overweight and obesity): −0.8 times/week (95% CI −1.5 to −0.1) French 2010 (SSB intake): +14 ml/day (P > 0.05) |
| F.4 Urban planning restrictions on new fast‐food outlets | SSB intake frequency (3 to 4 years) 11,821 inhabitants of 1 city and 1 county 1 CBA study ⊕⊝⊝⊝ VERY LOWd,g | Sturm 2015 (SSB intake frequency, measure or scale not reported): −0.9 (P > 0.05) |
| F.5 Restrictions to the number of stores selling SSB in remote communities | SSB sales (8 months) 3 stores in 1 remote community 1 ITS study ⊕⊝⊝⊝ VERY LOWd,g | Minaker 2016 (community‐wide SSB sales, model controlling for the summer peak): CAD −51/day (95% CI −166 to 65) |
| G Action across sectors | ||
| G.1 Trade and investment liberalisation in low‐ and middle‐income countries | SSB sales (4 years) 4 countries 2 controlled ITS studies ⊕⊝⊝⊝ VERY LOWd |
Baker 2016 (annual rate of change in volume sales of SSB per capita): −1.4 percentage points (95% CI −2.5 to −0.4) Baker 2016 (annual rate of change in volume sales of sugar from SSB per capita): −1.0 percentage points (95% CI −1.9 to −0.06) Baker 2016 (annual rate of change in volume sales of sports and energy drinks per capita): +0.3 percentage points (P > 0.05, SE 0.8) Schram 2015 (retail sales of SSB): +13 ml/per capita/day (95% CI 10 to 15) |
| G.2 Government food benefit programs with incentives for buying fruit and vegetables and restrictions on the purchase of SSB | SSB intake (3 to 12 months) 2274 adults and 18,207 children 3 RCTs with 5 comparisons ⊕⊕⊕⊝ MODERATEd | Collins 2016 WIC (intake of sugar from SSB, USD 60 versus no USD benefit/month):−5 g/day (95% CI −8 to −3) Collins 2016 WIC (intake of sugar from SSB, USD 60 versus USD 30 benefit/month): −1 g/day (95% CI −3 to 2) Collins 2016 WIC (intake of sugar from SSB, USD 30 versus no USD benefit/month): −5 g/day (95% CI −8 to −2) Harnack 2016 (SSB intake, incentives + restrictions): −180 ml/day (95% CI −338 to −22) Harnack 2016 (SSB purchases, incentives + restrictions):−0.3 USD/day (95% CI −0.5 to −0.2) Olsho 2016 (energy intake from SSB): −5 kcal/day/person (95% CI −21 to 11) Olsho 2016 (sugar intake from SSB): −1 g/day/person (95% CI −5 to 2) |
| G.3 Government food benefit programmes without incentives for buying fruit and vegetables and restrictions on the purchase of SSB |
SSB intake (3 to 8 months) 25,150 children and 2844 adults 1 RCT with 3 comparisons and 1 CBA study ⊕⊕⊝⊝ LOWd,g (Collins 2016 SNAP) ⊕⊝⊝⊝ VERY LOWd (Waehrer 2015) |
Collins 2016 SNAP (intake of sugar from SSB, USD 60 vs no USD benefit/month):−0.5 g/day (95% CI −2 to 1)
Collins 2016 SNAP (intake of sugar from SSB, USD 60 vs USD 30 benefit/month): +1 g/day (95% CI −1 to 3)
Collins 2016 SNAP (intake of sugar from SSB, USD 30 vs no USD benefit/month): −2 g/day (95% CI −4 to 1) Waehrer 2015 (SSB intake, median): +34 kcal/day (95% CI 7 to 60) |
| G.4 Multi‐component community campaigns focused on SSB | SSB sales (3 years) 32 supermarkets from 6 chains in 2 counties 1 controlled ITS study ⊕⊕⊕⊝ MODERATEc | Schwartz 2017 (SSB sales per product and store): −1.6 l/day (95% CI −2.0 to −1.2) (equivalent to a −20% decrease in the IG and a 0.8% increase in the CG) Schwartz 2017 (sports drinks sales per product and store):−0.4 l/day (95 CI −1.5 to 0.7) Schwartz 2017 (fruit drinks sales per product and store): −1.5 l/day (95% CI −2.0 to −0.9) |
| H Home‐based interventions | ||
| H.1 Improved access to low‐calorie beverages in the home environment | SSB intake (4 to 12 months) 1130 children, teenagers and adults 6 RCTs ⊕⊕⊕⊝ MODERATEd | Albala 2008, Anand 2007, Ebbeling 2006, Ebbeling 2012, Hernández‐Cordero 2014 (SSB intake, pooled effect estimate): −413 ml/day (95% CI −684 to −143) Tate 2012 (energy intake from beverages):−88 kcal/day (95% CI −124 to −51) |
| H.2 Provision of active video‐gaming equipment to teenagers |
SSB intake (10 months)
262 teenagers 1 RCT ⊕⊝⊝⊝ VERY LOWg,l |
Simons 2015 (share of participants consuming more than 1400 ml SSB a week): Decrease (OR 0.71, 95% CI 0.36 to 1.41) |
| CBA: Controlled‐before‐after study; CI: Confidence interval; ITS: interrupted‐time‐series study; NRCT: non‐randomised controlled trial; OR: odds ratio; RCT: randomised controlled trial; SSB: sugar‐sweetened beverages | ||
aThis 'Summary of Findings' table presents data on effects on direct and indirect measures of SSB intake only. We present data on the remaining primary outcomes (diet‐related anthropometric measures and adverse outcomes) in the additional 'Summary of Findings' tables displayed between the Results and the Discussion section. Data on secondary outcomes are presented narratively in the section on Effects of interventions.
b We assessed the certainty of evidence with GRADE. In GRADE, the certainty of evidence is rated as high, moderate, low or very low. Evidence from randomised controlled trials starts as high‐certainty, and evidence from non‐randomised studies starts as low‐certainty. From there, five factors that can decrease the certainty of evidence (risk of bias, indirectness, inconsistency, imprecision, and publication bias) and three factors can increase the certainty of evidence (magnitude of effect size, direction of plausible confounding, and dose‐response gradient). The certainty of evidence is understood in GRADE as the extent to which one can be confident that the true effect of an intervention lies on one side of a specified threshold, or within a chosen range (Hultcrantz 2017). For the present review, we defined this threshold as difference from the null. Our assessment of the certainty of evidence therefore refers to the existence and the direction of effects, and not to the exact effect sizes reported by individual studies. We report further details in the Methods section, and in the footnotes to the Summary of Findings tables 2‐9.
cUpgraded for magnitude of effect. dDowngraded for risk of bias. eDowngraded for indirectness. fVan de Gaar 2014 and French 2010 were cluster‐RCTs with four units, of which two were allocated by the flip of a coin to the intervention group. Given the small number of randomised units, randomisation may not have been sufficient to attain baseline comparability of the intervention and control groups. We therefore included the two studies as NRCTs in our evidence synthesis. gDowngraded for imprecision. hDowngraded by one level for risk of bias and by one level for imprecision (Da Costa 2014). iDowngraded by one level for risk of bias (Øverby 2012). jDowngraded for probability of publication bias. kUpgraded by one level for dose‐response gradient and magnitude of effect. lDowngraded for probability of publication bias.