Current evidence from large observational studies supports a link between sugar-sweetened beverage (SSB) consumption and the development of obesity in children and adults(Reference Malik, Pan and Willett1–Reference Bes-Rastrollo, Sayon-Orea and Ruiz-Canela3). Furthermore, 184 000 annual deaths worldwide were attributed to SSB consumption in 2010, mostly due to type 2 diabetes (72 %) or CVD (24 %), with 71 % of these deaths occurring in middle-income countries (MIC)(Reference Singh, Micha and Khatibzadeh4). Mexico, an MIC, implemented an SSB tax in 2014 and this fiscal policy was accompanied by a major mass communication strategy. Key messages were published on billboards or advertisements and posters in places such as metro stations, streets with significant foot traffic, and avenues where the soft drink industry advertised(5). Two years after the introduction of a tax on sugary drinks, households showed an average reduction of 7·6 % in the purchase of these drinks while households with lower incomes had a mean decrease of 11·7 %(Reference Colchero, Rivera-Dommarco and Popkin6).
The WHO has endorsed taxes on SSB, and in a report on fiscal policies for the prevention of non-communicable diseases, the WHO encouraged countries to increase taxation on SSB(7,8) . The organisation asserted that ‘imposing a tax on SSB is a potential strategy for increasing the price of these products and thus reducing their consumption’(9). A recent review indicates that in general, taxes decrease the consumption of taxed beverages by adults, though not for all types of beverages or all groups of consumers(Reference Cawley, Thow and Wen10).
Previously published systematic reviews, primarily based on studies from high-income countries (HIC), indicate that taxes on SSB may reduce the consumption of these products, although there is no evidence as to whether the effect would be similar in low-income countries and MIC(Reference Eyles, Ni Mhurchu and Nghiem11–Reference Thow, Annan and Mensah13). While some studies argue that the tax may work in similar ways in MIC and HIC(Reference Cabrera Escobar, Veerman and Tollman12), other studies have predicted that the tax may be more effective in MIC, given the evidence that consumers living in these countries are more responsive to price changes, since food costs are a primary determinant of consumption patterns, especially regarding non-essential products such as SSB(Reference Green, Cornelsen and Dangour14).
Although several systematic reviews have evaluated the impact of SSB taxes on consumption(Reference Alagiyawanna, Townsend and Mytton15–Reference Teng, Jones and Mizdrak17), few of them(Reference Cabrera Escobar, Veerman and Tollman12,Reference Nakhimovsky, Feigl and Avila18) have evaluated the impact of SSB taxes on overweight and obesity prevalence according to the income classifications of the countries. Furthermore, an updated review is required to take account of the recent growth in the number of SSB tax assessments from around the world.
Thus, this systematic review intends to answer the following research questions: (i) Is there evidence of an effect of implemented SSB tax policies on reducing consumption, purchase and sales? (ii) Is there evidence of an effect of implemented SSB taxation on reducing overweight and obesity prevalence? and (iii) Does the evidence of an effect of these differ according to the countries’ income classification?
Methods
Protocol and registration
The review was conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines(Reference Liberati, Altman and Tetzlaff19). The full study protocol has been registered on PROSPERO (International Prospective Register for Systematic Reviews, platform protocol number: CRD42020161612).
Primary exposures and outcomes
SSB taxes were the main exposure. SSB are beverages with added energetic sweeteners, such as sucrose, high-fructose corn syrup or fruit juice concentrates. These include non-alcoholic beverages, carbonates, fruit drinks, sports drinks, energy and vitamin water drinks, sweetened iced tea, and lemonade(7). An SSB tax was defined as a tax applied to at least one category of non-alcoholic beverages containing added sugars. Taxes could comprise excise taxes, import tariffs, sales taxes or any other taxes applied(Reference Heise, Katikireddi and Pega20). The main outcome was a change in overweight and obesity prevalence or change in body weight or BMI mean. We also assessed changes in the consumption of SSB, including any alteration in taxed beverage sales/purchases (as a proxy for consumption data), or dietary intake following the implementation of an SSB tax and these could be reported by volume, energies or consumption frequency.
Search strategy
Two reviewers conducted a blinded and independent literature search in December 2019. The following databases were searched: Cochrane Library, Embase, LILACS (via Virtual Health Library) and MEDLINE (via PubMed), and Web of Science. The search terms aimed to identify three domains: financial, nutritional and outcomes (Electronic Supplemental Material).
After the initial selection, reference lists from eligible studies and systematic reviews were searched for additional relevant studies. Peer-reviewed publications and grey literature (reports and self-published research) were included.
Eligibility criteria
Studies eligible for inclusion in this review were: (i) studies based on primary quantitative research, including modelling, non-experimental, quasi-experimental or experimental studies; (ii) full text published; (iii) written in English or Spanish; and (iv) published between January 2009 and December 2019. This review excluded other systematic reviews and meta-analyses as well as qualitative studies, case studies and reports, and letters to the editor. We included studies from all countries and average income of the country was classified according to World Bank definitions(21).
Study selection
After searching the databases, we used a free reference manager (Mendeley Desktop – Version 1.19.5) for article sorting, duplicate documents, reference organisation and search-time optimisation. Titles and abstracts were screened to identify relevant articles. Then, the reviewers, blinded and independently, identified and excluded the studies that did not meet the inclusion criteria. Finally, the full articles were obtained, and their eligibility was confirmed by full reading. Disagreements were solved by consensus between the two reviewers; there was no need for arbitration by a third reviewer.
Assessment of study quality
Each eligible study was assessed for study quality using a critical appraisal tool (Electronic Supplemental Material) established according to previously published reviews of food and beverage pricing studies(Reference Nakhimovsky, Feigl and Avila18,Reference Thow, Downs and Jan22,Reference Backholer, Sarink and Beauchamp23) .
The two reviewers applied the above quality criteria for all eligible studies and rated each of them on seven criteria, establishing for the presence of ‘1 point’ or absence of ‘0 points’ for each item. All articles that achieved a score equal to or higher than 4 points in these quality criteria were selected for this systematic review.
Data extraction
Using a standardised electronic form, two reviewers independently extracted the following data from the studies:
-
General information: authors, year of publication, country, income classification, study design and sample size;
-
Fiscal outcomes: tax rate, taxed products, year of policy sanction, consumption, purchase, sale and major findings;
-
Nutritional outcomes: overweight and obesity prevalence, changes in BMI, change in body weight, and major findings.
Synthesis of the results
The effect of SSB taxes on outcomes of interest (change in overweight and obesity prevalence, body weight and BMI, consumption, sales, and purchase) were reported as difference, percentage change and tax elasticity. Tax elasticity is the percentage change in consumption for a 1 % change in tax.
Results
A total of 8338 articles were identified after the database search plus 11 additional records from the reference lists of eligible studies. After excluding 552 duplicated records, 2125 published before January 2009, 122 not written in English or Spanish (German: 34 studies; French: 20 studies; Chinese: 19 studies; Italian: 13 studies; Russian: 12 studies; Polish: 7 studies; Dutch: 5 studies; Japanese: 4 countries; Norwegian: 2 studies; and Icelandic, Danish, Bulgarian, Hungarian, Slovenian, Swiss: 1 study in each language) and 5145 due to study type, 405 records were screened by title, and 98 were selected for full-text reading. Of those, twenty-one were eligible for the review and eighteen met quality criteria. Among the eighteen articles included in this systematic review, twelve are modelling studies and six are studies that evaluated the impact of implemented real taxes. The detailed selection process is shown in Fig. 1, and the quality checklist applied to the included studies is shown in Table 1.
Fig. 1 Flow chart of the selection of studies evaluating the impact of SSB tax on consumption, purchase, sale, overweight and obesity, published between 2009 and 2019. SSB, sugar-sweetened beverages
Table 1 Quality checklist for included studies
Q1. Does the study consist in a prospective evaluation of observed behaviour within the same population?
Q2. Do price and consumption data come from the same population?
Q3. Does the study assess an actual tax or subsidy rather than hypothetical measures?
Q4. Does the data include all SSB consumed?
Q5. Does the study report the effect of SSB intake on overweigh or obesity within the same population?
Q6. Does the study consider potential substitution to other products?
Q7. Does the study consider a long-run input data across time with sufficient variation in prices used to estimate price elasticities? (For experimental studies: a period of at least 1 month; for studies using existing data sets on SSB price: intervals no less than 2 months apart for at least 12 months).
The quality of studies was analysed using a tool specific for outcomes of the SSB tax and the scores ranged from 4 to 5 in HIC, upper-income countries and MIC. We considered that the quality was similar among studies. Item 1 was observed only in two studies from HIC, but the objective was not to assess an SSB tax on overweight/obesity. A prospective design is better than a cross-sectional one to answer this objective. Items 2 (Q2) and 5 (Q5) were met in sixteen studies and represented that the price and consumption data of SSB or effect of SSB intake on overweight/obesity was from the same population. We considered that these aspects must be fulfilled for our analysis. Item 3 was scored in nine studies, independently of income classification. However, this is probably not a relevant limitation as the tax rate had not yet been defined in the country. Item 4 was scored in twelve studies, including upper-income countries and MIC. We assumed that in countries that included only one type of SSB, for example soft drinks, that the intake of this beverage was higher than others or the price elasticity for them was not available. Item 7 was scored in nine countries, and we considered that the effect of an SSB tax on obesity consumption could be observed(Reference Sturm, Powell and Chriqui24–Reference Powell, Chriqui and Chaloupka27) or could be higher than we observed(Reference Briggs, Mytton and Kehlbacher28–Reference Briggs, Mytton and Madden30), if a minimum interval between the implementation of an SSB tax and effect evaluation were considered.
Table 2 describes the sixteen studies found in this systematic review that evaluated the impact of SSB taxes on the consumption, purchase or sales of these products, and 75 % were developed in HIC. Sales and purchase were evaluated in five studies(Reference Sturm, Powell and Chriqui24,Reference Nakamura, Mirelman and Cuadrado25,Reference Briggs, Mytton and Kehlbacher28,Reference Silver, Ng and Ryan-Ibarra31,Reference Alvarado, Unwin and Sharp32) . Sales decrease of 8·6 ml/capita/week after applying a 10 % ad valorem tax(Reference Alvarado, Unwin and Sharp32) in Barbados (time series analysis) and 9·6 % following a US$0·67¢/oz price increase(Reference Silver, Ng and Ryan-Ibarra31) in California (comparison between pre-tax and first year post-tax) were found. Purchase decrease of 15 % for a 20 % sales tax in the United Kingdom(Reference Briggs, Mytton and Kehlbacher28) (theoretical simulation models) and 22 % for 1·6 % price increase in Chile(Reference Nakamura, Mirelman and Cuadrado25) (comparison between pre-tax and post-tax) were found, but no statistically significant association was found in children’s consumption of SSB for a 4·2 % (mean) state-level sales tax in grocery stores in the USA (comparison between pre-tax and post-tax)(Reference Sturm, Powell and Chriqui24).
Table 2 Summary of study findings evaluating the potential impact of sugar-sweetened beverages (SSB) taxes on dietary intake, purchase or sales
* ‘Impuesto Adicional a las Bebidas Analcohólicas’ or additional tax on nonalcoholic drinks.
† 3 % tax for soft drinks containing low sugar levels, 18 % tax for beverages above an added sugar concentration of at least 6·25 g/100 ml and 10 % tax for those below this threshold.
‡ Resulting in a percentage increase in price, as opposed to an excise tax that varies with the content of the product.
According to three references, consumption was reduced by approximately 13 % (20 % added tax)(Reference Veerman, Sacks and Antonopoulos33), 20 % ($0·01 cent per fluid ounce excise tax)(Reference Long, Gortmaker and Ward34) and 40 % ($0·015 cents per fluid ounce excise tax)(Reference Zhong, Auchincloss and Lee35), and all studies applied theoretical simulation models. Energy intake was evaluated in most of the studies and showed a reduction of 8·8 to 69·9 kJ/d/person(Reference Briggs, Mytton and Kehlbacher28,Reference Silver, Ng and Ryan-Ibarra31,Reference Veerman, Sacks and Antonopoulos33) , of 25 kJ for each one percentage point increase in the soft drink tax(Reference Fletcher, Frisvold and Tefft26), and of 150 kJ/d(Reference Lin, Smith and Lee36). These studies applied theoretical simulation models except for Silver et al.(Reference Silver, Ng and Ryan-Ibarra31)
Three studies based on theoretical simulation models examined the effect of taxes from theoretical simulation models in upper-middle-income countries, and all of them showed an association between taxation and SSB purchase or energy intake. Reductions of 21·62 and 43·23 ml/person/d were seen for taxes of 10 and 20 %, respectively(Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37). An average reduction of 7·6 % in purchases(Reference Colchero, Rivera-Dommarco and Popkin6) and a reduction of energy intake by 36 kJ/d(Reference Veerman, Sacks and Antonopoulos33) were also found (Table 2).
Nine studies evaluated taxes applied to carbonated soft drinks or SSB(Reference Colchero, Rivera-Dommarco and Popkin6,Reference Sturm, Powell and Chriqui24–Reference Fletcher, Frisvold and Tefft26,Reference Briggs, Mytton and Kehlbacher28,Reference Briggs, Mytton and Madden30,Reference Silver, Ng and Ryan-Ibarra31,Reference Long, Gortmaker and Ward34,Reference Fletcher, Frisvold and Tefft38) and other studies assessed taxes applied to a broad definition of SSB (flavoured water, energy drinks and fruit juice drinks). Pan American Health Organisation recommends including all kinds of beverages with added sugar to prevent unhealthy substitutions and offering healthy options to replace SSB(5). The main outcome we observed in the study was the effect of an SSB tax on the consumption of these beverages (eleven studies), only two studies evaluated SSB sales(Reference Nakamura, Mirelman and Cuadrado25,Reference Silver, Ng and Ryan-Ibarra31) and two studies evaluated SSB purchases(Reference Colchero, Rivera-Dommarco and Popkin6,Reference Nakamura, Mirelman and Cuadrado25) . The main kind of SSB tax was excise tax expressed as a percentage increase in the final product price (10 to 20 %) and five countries applied a specific tax based on the weight of the SSB (the level ranged from $0·010 to $0·015/ounce of SSB). One study(Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37) based on theoretical simulation models compared excise taxes of 10 and 20 % and the decrease in SSB consumption was twice as high for the 20 % excise tax. Only two studies(Reference Zhong, Auchincloss and Lee35,Reference Basu, Vellakkal and Agrawal39) showed that applying taxes resulted in a decreased consumption of SSB and an increased consumption of water, milk, tea or fruit juice (cross-price elasticity). One of them was based on theoretical simulation models35 and the other evaluated the impact of an implemented tax(Reference Basu, Vellakkal and Agrawal39).
The effectiveness of SSB taxation in reducing the prevalence of overweight and obesity is shown in Table 3. The reduction in obesity prevalence varied from 0·99 % to 2·7 % points(Reference Bes-Rastrollo, Sayon-Orea and Ruiz-Canela3,Reference Briggs, Mytton and Madden30,Reference Veerman, Sacks and Antonopoulos33,Reference Long, Gortmaker and Ward34) and the reduction in overweight prevalence was 0·7 %(Reference Briggs, Mytton and Madden30), all of them were theoretical simulation models. Decreases in weight and BMI were 0·97 kg(Reference Lin, Smith and Lee36) and 0·08 kg/m2(Reference Long, Gortmaker and Ward34), respectively (theoretical simulation models) (Table 3). Two studies(Reference Manyema, Veerman and Chola29,Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37) from upper-middle-income countries using theoretical simulation models showed a positive effect of an SSB tax in decreasing obesity prevalence. Reductions of 2·5 and 5·3 % in obesity prevalence were reported for a 10 and 20 % taxation of SSB(Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37), respectively, and 3·8/2·4 % decreases in obesity prevalence were seen in men/women(Reference Manyema, Veerman and Chola29). The effect of an SSB tax on overweight and obesity prevalence identified from theoretical simulation models was higher in young individuals(Reference Briggs, Mytton and Kehlbacher28–Reference Briggs, Mytton and Madden30,Reference Long, Gortmaker and Ward34,Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37,Reference Basu, Vellakkal and Agrawal39) .
Table 3 Summary of study findings evaluating the potential impact of sugar-sweetened beverages (SSB) taxes on overweight and obesity outcomes
In five studies, there was no statistically significant association between taxing SSB and changes in BMI, or overweight/obesity(Reference Sturm, Powell and Chriqui24,Reference Nakamura, Mirelman and Cuadrado25,Reference Powell, Chriqui and Chaloupka27,Reference Fletcher, Frisvold and Tefft38,Reference Fletcher, Frisvold and Tefft40) . Most of these studies refer to modelling studies except for the Sturm et al. (Reference Sturm, Powell and Chriqui24), Powell et al. (Reference Powell, Chriqui and Chaloupka27) and Nakamura et al. (Reference Nakamura, Mirelman and Cuadrado25) studies. Sturm et al. (Reference Sturm, Powell and Chriqui24), using data on state sales taxes for soda and individual-level data on children, showed a non-significant association between an SSB tax and energy intake for children. However, the authors highlighted that an SSB tax could be more effective for at-risk groups such as overweight children, families with lower socio-economic status and African-American children, and for SSB sold in the school environment(Reference Sturm, Powell and Chriqui24).
In studies that applied an SSB tax of 1 or 3 % from theoretical simulation models(Reference Fletcher, Frisvold and Tefft26,Reference Fletcher, Frisvold and Tefft40) , no effect on overweight and obesity prevalence was observed. A limitation of these two studies is that SSB consumption information was taken from the National Health and Nutrition Examination Survey (population level) and the SSB tax was a different state-specific-tax, and the information on price and consumption was from a different population. The study developed by Powell et al. (Reference Powell, Chriqui and Chaloupka27) assessed the impact of applied taxes of 4·25 % for grocery stores and 4·51 % for vending machines which was insufficient to impact on BMI values. However, the authors highlighted a weak association between increased vending machine tax rates (contextual level) and BMI for overweight adolescents (individual level)(Reference Powell, Chriqui and Chaloupka27).
In Chile, a 13–18 % excise tax was applied to beverages with equal or higher than 6·25 g of sugar/100 ml and a parallel reduction of 3 % in the price of beverages with lower amounts of sugar. The purchase of SSB decreased after implementation of the SSB tax, but no effect of SSB tax on nutritional status was observed (comparison between pre-tax and post-tax)(Reference Nakamura, Mirelman and Cuadrado25). In general, the results of the studies analysed show that a higher taxation (20 %) has a greater impact on sales, and purchase, and a decrease in overweight and obesity compared to more modest taxes (10 %).
Discussion
This systematic review covered two main aspects of SSB taxes. The first was to quantify the effect of SSB taxes on consumption, purchase or sales, and the second was to evaluate the effect of SSB taxes on body weight, BMI or overweight/obesity prevalence. In all, eighteen studies were selected; sixteen of which answered the first objective and thirteen the second. Most studies were conducted in countries with a high-income level, especially in the USA (seven studies). The effect of SSB taxes on overweight or obesity prevalence was estimated mainly from modelling studies (twelve of eighteen studies), cross-sectional data, studies conducted with adults and those conducted before the implementation of an SSB tax in the country or state (US studies) as a fiscal policy.
In this review, all studies, except two(Reference Sturm, Powell and Chriqui24,Reference Fletcher, Frisvold and Tefft38) , showed a positive effect of SSB taxation, that is, consumption, sales and purchases decreased when the price of SSB was increased in MIC, upper-income countries and HIC. Sales were assessed only in two HIC. The most common unit of assessment SSB intake was expressed as energy intake reduction (kJ/d) and ranged from 8·8 kJ to 150·6 kJ in HIC, and 36 kJ/d in one upper-income country. There was a dose–response effect for taxes of 10 and 20 % (decreased intake from 21·6 ml to 42·2 ml, respectively). The effect of SSB taxes on the estimated amount of energetic reduction depends on SSB intake as a percentage of total energy intake, that is, in countries where intake is high, the effect will be higher. For example, an SSB tax of 10 % can decrease 75·7 kJ/d/person in the USA(Reference Lin, Smith and Lee36); 35 kJ/d/person in Mexico and only 8·8 kJ/d/person in Ireland(Reference Briggs, Mytton and Madden30). The demand of SSB is sensitive to changes in price and therefore, an SSB tax affects demand(Reference Guerrero-López, Unar-Munguía and Colchero41). For countries with higher price elasticity, the effect of an SSB tax will be higher. Barrientos-Gutierrez et al. (2017)(Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37) highlighted that it is relevant to analyse the cross-price elasticity of an SSB tax, but it depends on the availability of these data in countries. In our systematic review, only two studies included cross-price elasticity, and both showed that there was an increased consumption of bottled water(Reference Manyema, Veerman and Chola29) and milk, fruit juice and tea(Reference Basu, Vellakkal and Agrawal39) after SSB tax implementation. A systematic review developed in MIC showed that the tax applied to soft drinks had a greater effect than sweetened fruit juices (taxation of 10 % in price contributed to a decrease of 5 to 39 kJ per capita/d in soft drink consumption). The authors observed a concurrent increase in milk consumption(Reference Nakhimovsky, Feigl and Avila18).
We identified thirteen studies that evaluated the effect of SSB taxes on BMI or overweight and obesity prevalence, and only three studies were conducted in upper-middle-income countries (Mexico, South Africa) and MIC (India). We observed a decrease in obesity prevalence in six HIC (60 %) and in all upper-income countries and MIC (100 %). All these studies were of cross-sectional design from national surveys, one modelling study included adults and similar SSB categories, the excise tax applied was 10 or 20 % and the own-price elasticity ranged from –0·8 to –1·29. The percentage of obesity changed may be higher in MIC and upper-income countries, as it ranged from 2·54 to 5·9 % in these countries v. 0·99 to 2·7 % in HIC. In all countries except Chile, the studies were conducted before an SSB tax was adopted as a fiscal policy to decrease SSB consumption(Reference Nakamura, Mirelman and Cuadrado25). Australia(Reference Veerman, Sacks and Antonopoulos33) was the only country where an SSB tax has not yet been implemented as a fiscal policy. According to Sainsbury et al. (Reference Sainsbury, Magnusson and Thow42), difficulty in implementing a fiscal policy is due to influence from industry, fragmented advocacy efforts, political opposition to paternalistic policies, conflicting political agendas and inadequate pressure for change from civil society.
The models used to estimate weight change as a function of energetic intake change after an SSB tax implementation in the studies were based on a dynamic model and may contribute to avoid overestimation of the effect of SSB taxation on BMI values. These equations consider a new ‘steady state’ of body weight that is achieved after the change in energetic intake and is considered better than the static model(Reference Lin, Smith and Lee36). Four studies(Reference Briggs, Mytton and Kehlbacher28,Reference Long, Gortmaker and Ward34,Reference Barrientos-Gutierrez, Zepeda-Tello and Rodrigues37,Reference Basu, Vellakkal and Agrawal39) included a cross-elasticity demand together with own elasticity of SSB and observed the replacement of SSB by other beverages (milk, fruit juice, water and diet soft drinks). Cross-elasticity analysis would be relevant to prevent overestimation of the SSB tax effect on consumption, sales or purchases.
This review does not include a meta-analysis due to the heterogeneity of methods for estimating change in overweight and obesity, SSB consumption, study design, and difficulties in controlling for income, population and other country-specific characteristics. The strengths of our study are the application of a quality checklist to include studies that specifically focus on SSB taxes, the selection of studies from different income levels and addressing the effects of SSB taxes on sales, consumption, purchase and overweight/obesity prevalence in the same review.
The limitations of our study are focused on the lack of some relevant data from selected studies to assess their quality. We applied a critical appraisal tool to assess seven topics on the quality of studies before inclusion in our systematic review. The first item on our checklist (Q1) was only met by two studies(Reference Silver, Ng and Ryan-Ibarra31,Reference Zhong, Auchincloss and Lee35) and these studies did not evaluate the effect of an SSB tax on overweight/obesity prevalence. Another relevant question with lower scores was question 7 (Q7: minimum interval between the implementation of an SSB tax and effect evaluation), as discussed earlier. We also highlighted the absence of an accurate definition of SSB in some studies. This term is broad and sometimes refers to soda, or soft drinks, or carbonated soft drinks.
Data about own-price and cross-price elasticities were not available in some countries and it was necessary to apply data from other countries. The studies highlighted difficulties in selecting the most appropriate model to explain weight loss and to estimate the effect of reduced SSB consumption on weight loss. The lack of information about other health policies that were carried out in these countries in addition to the SSB tax also might compromise the evaluation of the impact from fiscal policy. In Mexico, the government launched a mass-media campaign on healthy habits as well as an SSB tax implementation(5,7) . Thow et al. (Reference Thow, Downs and Jan22) recommended developing prospective studies and estimating the effect of SSB taxes with other interventions. Another aspect to be considered is that weight is associated with many factors of different nature so that isolated interventions would possibly not have a relevant impact on its reduction, requiring the implementation of broader intervention packages. Finally, we did not find any studies from low-income countries. One reason for this could be that per capita consumption is still low (0·3 portion/d), being lower than that of all the other regions(Reference Nugent43). Also, in low-income countries, there are other serious health problems. We suggested that is relevant to estimate the price elasticity of SSB in low-income countries and MIC, because types of taxes may be different, and the characteristics of SSB sales and consumption are also different.
The results of our systematic review showed that an SSB tax could be an effective fiscal policy to decrease the purchase and consumption of SSB and reduce overweight/obesity prevalence. The effect of an SSB tax would be more powerful if the tax were higher, were specific for beverage volume and covered all kinds of SSB. Future longitudinal studies that include an SSB tax already implemented in the country as a fiscal policy and data about individual consumption of SSB are needed for an accurate estimate of overweight/obesity prevalence.
Acknowledgements
Acknowledgements: None. Financial support: This study received funding from Conselho Nacional de Desenvolvimento Científico e Tecnológico 442630/2019-0. Conflict of interest: There are no conflicts of interest. Authorship: All authors conceived the review. C.C.E., A.E.M.R. and L.B.N wrote the first draft of the manuscript. C.C.E., S.S.B. and A.I contributed to the development of search strategy. A.I., C.C.E., S.S.B., A.E.M.R. and L.B.N contributed to the writing of the manuscript. A.I. and S.S.B contributed to the selection of studies. A.I. and S.S.B contributed to the extraction of data. A.I. and S.S.B contributed substantially to the methods for addressing quality of studies in the review. All authors read and approved the final version of the manuscript. Ethics of human subject participation: Not applicable.
Supplementary material
For supplementary material accompanying this paper visit https://doi.org/10.1017/S1368980021002901
