Table 1 Summary of included systematic reviews with data pertaining to Question 1^a

Itria et al. [15]. AMSTAR Rating: High

 To assess the effect of implemented SSB tax policies on reducing consumption, purchase and sales, overweight and obesity prevalence

Any type of tax

Studies included modelling, non-experimental quasi-experimental or experimental

Peer-reviewed and grey literature

Australia; Barbados; Chile;

Ireland; Mexico;

South Africa; UK;

USA

(African, European, Western Pacific, The Americas)

Narrative

Outcomes included changes in SSB consumption including purchase/sales as proxy for consumption. (See data from individual studies in Additional file 6)

16 papers: [42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57]

Out of 7 elements of quality assessment, all included studies achieved scores of 4–5

Sobhani et al. [30] AMSTAR Rating: Moderate

 To assess evidence on the impacts of SSB taxation on purchase and consumption of SSBs

Excise tax/sales tax

Empirical data excluded modelling studies. Included quasi-experimental (pre- post comparisons) and RCT designs

Mexico

USA

Holland

(European, The Americas)

Narrative

Outcomes included consumption, purchase, or sales of SSB (See data from individual studies Additional file 6)

5 studies with data relevant to the question of this review: [44, 58,59,60,61]

Out of 7 elements of quality assessment, all included studies achieved scores of 4–5

Teng et al. [27] AMSTAR Rating: High

 To meta-analyse data from studies measuring impact of real-world SSB taxes on SSB sales and intake before and after the tax, or in a taxed compared with an untaxed jurisdiction

Studies with data on the impact of implementation of real world SSB taxes

All types of SSB taxes (excise, sales, volumetric, ad valorem, sugars content-based taxes

All were HIC except Mexico (high-middle)

Chile

Finland

France

Hungary

Mexico

Spain

USA

(European, The Americas)

Meta-analysis: the summary measure was a RR scaled for 10% ad-valorem tax (AVEs were applied to volumetric tax). Sub-analysis by tax type

Outcomes included purchase or dietary intake

Meta‐analysis showed 10% SSB tax decreased consumption by 10.0% (95% CI: − 5.0, -14.7%), based on pre–post intervention comparisons and/or comparisons to an untaxed control jurisdiction. This corresponded to PED − 1.00 (95% CI: − 0.50, − 1.47)

By tax type: 2.3% (95% CI: -11.2, 7.4%) decline for ad valorem tax;

a 10.2% decline (95% CI: -4.1, -15.9%) for volumetric;

14% (95% CI: -7.5, -20.1%) for with a sugar concentration threshold (NS)

By age group: all 10.2%, adults 6.4%, children 7.7% p = 0.91

By SES (NS)

Papers included in meta-analysis: [42, 44, 47, 51, 53, 56, 58, 60, 62,63,64,65,66,67,68]

Risk of bias assessment classified 8/22 studies high, 6/22 moderate, and 8/22 low quality

Bergallo et al. [31] AMSTAR Rating: Moderate

 To assess the available evidence on impact of SSB taxation (and other regulatory strategies to reduce SSB) in Latin American countries

Excise tax, and VAT

Included naturalistic studies (experimental and modelling excluded)

Peer-reviewed studies only

Mexico, Barbados

(The Americas)

Narrative

Outcomes included consumption, purchase, or sales of SSB (See data from individual studies Additional file 6)

4 papers: [53, 58, 69, 70]

No appraisal of quality of included studies

Redondo et al. [32] AMSTAR Rating: Moderate

 To synthesize existing evidence related to the impact of SSB taxes on consumption, purchase, or sales

All types of taxes

5 Naturalistic

Studies (excluded simulations/modelling studies

USA

Mexico

(The Americas)

Narrative

Outcomes included SSB sale, purchasing behaviour. (See data from individual studies Additional file 6)

5 papers: [44, 58,59,60, 62]

Quality appraisal addressed 58 elements. Elements not met ranged from 2–6. Elements met ranged from 33–48,

Afshin et al. [28]. AMSTAR Rating: High

 To quantify the prospective effect of changes in food prices on dietary consumption

Included intervention and PCS on impact of change in price on consumption and adiposity

Excluded modelling studies, cross-sectional studies, and laboratory experiments (hypothetical

situations)

USA

(The Americas)

Study-specific effect sizes were pooled using inverse-variance-weighted random-effect models

Each 10% price increase reduced SSB intake by 7% (95% CI: -3.0, -10.0%)

4 papers: [71,72,73,74]

Quality scored out of 5 ranged from 3–5

Nakhimovsky et al. [29] AMSTAR rating: High

 To assess the effectiveness of SSB tax in MIC in comparison to HIC. To assess: (1) prices increase after introduction of excise tax; (2) impact on EI and differences across SES groups

Middle-income countries

Any tax or price change on SSB consumption (and preferable body weight outcomes). All SSB categories included

Modelling, non-experimental, quasi-experimental or experimental studies

Brazil, Ecuador India Mexico Peru

South Africa

(African, The Americas)

Standardised estimates for change in SSB intake to kJ/person/day for a 10% price increase

Analysis showed OPE (range: -0.6- to -1.2) equivalent to a 6.0% to 12.0% reduction in consumption with 10% tax. With 10% tax EI reduction of 18.0 (range: 5.0 to 39.0) KJ/ person/day (~ 1.07 (range: 0.3 to 2.3) g sugars /person/day). Assumed linearity between % change in price and intake

Reduction in SSB purchase higher (9.1%) in lowest SES tertile than highest (5.5%). OPEs higher among lower income groups (no meta-analysis of these data: see individual studies in Additional file 6)

Nine studies, (three quasi-experimental [59, 75, 76]

Six

observational & modelling [55, 57, 59, 77,78,79]

Assessed studies against 6 criteria checklist. Mode = 4. Range 1–5 out of 6. Used hierarchy of study design to inform findings

Backholer et al. [33] AMSTAR Rating: Moderate

 To clarify the differential impact(s) of SSB taxes on SSB purchase, intake, and body weight outcomes by SES

Studies that reported the impact of change in SSB price on purchase, intake, or EI and/or body weight related outcomes according to any marker of SES in HIC

USA

UK

Ireland

New Zealand

(European, Western Pacific, The Americas)

Narrative

Outcomes included SSB intake, EI, and body weight outcomes. (See data from individual studies Additional file 6)

8 papers with relevant data: [45, 46, 51, 52, 80,81,82]

Quality of included studies was not appraised

Niebylski et al. [39] AMSTAR Rating: Low

 To evaluate published evidence to assess the effect of healthier food/beverage subsidies and less food/beverage taxation on diet

Included studies, reviews, and/or predictive models for adults and children in Western Europe, Canada, US, Australia, and New Zealand. Articles assessing tax (any type) effect on nutrition-related health outcomes and consumption of foods including SSB

Canada

USA

(The Americas)

Narrative

Outcomes included intake of healthier vs. less health (none core) foods, including SSB (See data from individual studies Additional file 6)

15 papers: [29, 37, 46, 52, 71, 73, 78, 83,84,85,86,87,88,89,90]

GRADE method used to assess quality of body of evidence

Thow et al. [34] AMSTAR Rating: Moderate

 To assess the effectiveness of food tax and subsidies policies on consumption

Studies with empirical data on impact of food or nutrient tax (any type) on subsidy on consumption. Modelling and stated preference studies included

USA, Norway, Brazil, France

(European, The Americas)

Narrative reporting the range of effect across studies

SSB taxes ranged from 5 to 30%. All showed a reduction in SSB consumption range: 5% to 48%. Overall change in consumption was proportional to the taxes applied

SES: four modelled studies from Brazil, Finland and US found higher price sensitivity for low-income households

(See data, including for cross price elasticity from individual studies Additional file 6)

Sixteen studies: [38, 42, 51, 52, 56, 73, 78, 81, 83, 87,88,89,90,91,92,93]

Assessed quality with reference to:

1) Is study prospective

2) Data includes all SSB

3) Price and consumption data from same population 4) Considers potential substitution

5) Based on individual food consumption- yes

6) Study assesses an actual tax – no

Scores ranged 1–4/6:

1 study 1/6,

6 studies 2/6,

6 studies 3/6, 1 study 4/6)

Cabrera Escobar et al. [35] AMSTAR rating: Moderate

 To evaluate the published evidence for impact of SSB taxes or price increases, on consumption levels and obesity, overweight and BMI

Studies with original evidence on the quantitative impact of SSB price changes on consumption or on weight change or BMI

Articles in English 2000–2013

USA, Mexico, Brazil, France

(European, The Americas)

Meta-analysis on OPE and cross-PED using random effects model

All the results show negative elasticity; OPE -1.3 (95% CI: -1.09, -1.51)

Mexico -1.09 (SE 0.20)

France -2.21 (SE 0.13)

Brazil -0.85 (SE 0.43)

Four studies (3 USA, 1 Mexico) reported cross-PED showing increased SSB price increased demand for other beverages, fruit juice 0.39 (95% CI: 0.01, 0.77), and milk 0.13 (95% CI: -0.09, 0.34), and reduced demand for diet drinks − 0.42 (95% CI: -0.63, -1.23)

The evidence from LMICs consistent with HIC countries

9 Studies: [42, 52, 56, 77, 78, 83, 87, 88, 91, 94]

Quality of studies not assessed

Maniadakis et al. [36] AMSTAR Rating: Moderate

 To assess the impact of tax policies and price increases on SSB consumption, EI and weight outcomes

Included were original studies in the four categories of: existing data; experiments; survey; and observational studies; – on the association between SSBs prices and taxes and corresponding intake, EI or obesity-related outcomes

USA, UK, Norway, Italy, Denmark, Germany, France, the Netherlands, Mexico, Brazil, Taiwan, Singapore, and Australia

(European, SE Asian, Western Pacific, The Americas)

Narrative

Studies (n = 9) with data on prices/taxes and PED showed PED ranged from -0.5 to -1.6 with most < 1.0. PED was more regressive towards the lower SES

2 studies with data on SSB tax and EI showed a 10% increase in price/tax reduced EI by up to 50 kcal/day or. 450/month (equivalent to a reduction in sugars of ~ 12.5 g/d and 112.5 g/month)

Included in PED synthesis [71, 85, 95,96,97,98,99,100,101]

Reported narrative: 27, 50, 51, 55, 73, 79, 80, 83, 90, 94, 98, 106

Quality of studies not assessed

Powell et al. [37] AMSTAR Rating: Moderate

 To assess PED for SSBs, fast food, and fruits & vegetables, and direct associations of prices/taxes with body weight outcomes

Included studies with US data, original quantitative data on effect of price/tax on consumption or weight change. Studies that assessed demand for product category (not brand) excluded intervention studies

USA

(The Americas)

Derived an overall mean estimate of PED of SSB from available SSB estimates for all SSB/sub-categories

Each PED estimate was weighted by its relative consumption share of SSBs based on EI data from individual dietary data for those aged 2 + from the 2007–2008 NHANES

Based on 12 available price elasticity studies, the overall OPE was -1.21 (95% CI: -0.71, -3.87) implying that a 10% tax would reduce consumption by 12% and a 20% tax by 24%

11 studies included [51, 52, 73, 81, 83, 87, 88, 96, 101,102,103]

Quality of studies not assessed

Andreyeva et al. [38] AMSTAR Rating: Moderate

 To estimate effect of price change on consumer demand for major commodity foods, and to explored how altering SSB price alters consumption

Studies with data on SSB reporting on PED

USA

(The Americas)

Mean PED estimates and variations in estimates by study design

Average PED 0.79 (95% CI: 0.33, 1.24), range 0.13 to 3.18) based on 14 estimates. A10% tax on SSB reduced consumption by 8–10%

14 studies with data on SSB available from authors