Sugar-sweetened soft drinks deliver liquid calories that do not satisfy hunger and therefore promote weight gain in children and adults.
Original statement
المشروبات الغازية — السكرية منها والدايت — تُحمِّل الجسم سعرات سائلة لا تُشبع
verbatim according to source. Translation by the project; the Arabic wording is authoritative.
Translation
Sugar-sweetened soft drinks deliver liquid calories that do not satisfy hunger and therefore promote weight gain in children and adults.
Origin
- Source status
- Archive editors
- Attribution
- Archive editors (study selection and narrative are the archive's, not necessarily Elawady's)
- Origin
- https://eltayebat.com/rule/soda/
- Attribution uncertainty
- The archive homepage states that the forbidden-food section is supported by research 'independently of Dr. Diaa's words'. No evidence was found that Elawady himself cited these studies.
- Claimed mechanism
Weak dietary compensation for liquid calories.
- Claimed health effect
Weight gain, obesity.
- Population concerned
Children and adults
- Geographic restriction by Elawady
none found
Scientific assessment
Mostly true – Correct at its core, but with limitations or minor inaccuracies.
Certainty of evidence: moderate – Further research may still affect the assessment.
1. What was claimed?
Liquid sugar calories do not satisfy and cause weight gain.
2. What is demonstrably correct?
Malik 2013 (A-SOD-20): cohorts show weight gain per daily serving and RCTs in adults adding SSB showed +0.85 kg (95% CI 0.50-1.20). Te Morenga 2012 (I-SUG-01): reducing sugars reduced weight by 0.80 kg in ad libitum trials. DiMeglio 2000 (A-SOD-21): in 15 people, energy from soda was not compensated whereas energy from jelly beans was.
3. What is wrong or exaggerated?
The summary sentence groups diet sodas with sugary ones under 'liquid calories'; diet sodas provide almost no energy. Te Morenga found no weight change when sugars were exchanged isoenergetically for other carbohydrates, so the effect works through extra energy intake rather than a unique property of sugar.
4. What remains unclear?
Long-term size of the satiety effect; the DiMeglio trial is very small.
5. Which studies support it?
A-SOD-04, A-SOD-10, A-SOD-20, A-SOD-21, A-SOD-23; I-SUG-01; I-SUG-05.
6. Which studies contradict it?
Isoenergetic exchange trials (I-SUG-01) show no weight effect when total calories are held constant.
7. What applies to Germany and Europe?
The WHO guideline (2015) on free sugars and EFSA's 2022 opinion (intake of added and free sugars 'as low as possible') apply in Germany and the EU; the European EPIC cohort (A-SOD-17) found associations in 10 European countries including Germany.
8. What is known about Egypt?
Sugar-sweetened soft drinks are widely sold in Egypt. No Egypt-specific study was reviewed for this claim; the physiological mechanisms and cohort findings are not known to differ by country, but absolute risks depend on local consumption and disease rates.
9. What practical conclusion is appropriate?
Replacing sugary drinks with water is a reasonable weight-management step.
Assessment details
- Rationale for certainty of evidence
RCTs in adults (adding SSB increases weight) and in children (reducing SSB reduces BMI gain), plus consistent cohorts; WHO rates the evidence for free sugars and body weight as basis for a strong recommendation. Mechanistic evidence on poor satiety of liquid calories rests on few small trials.
- Source quality
Archive studies are relevant and correctly summarised; the claim that diet sodas also 'load liquid calories' is incorrect (diet sodas are essentially calorie-free).
- Regional transferability
High.
- Individual exceptions
Effect sizes per serving are modest (about 0.1-0.2 kg/year per daily serving in cohorts); large differences arise with high habitual intake.
- Regional scope of the claim
- general (biological/medical)
- Last reviewed
- 26 October 2026 provisional
Scientific evidence in detail
Scientifically well supported – The claim agrees with consistent, high-quality evidence.
Research question: Do sugar-sweetened soft drinks promote weight gain in children and adults, partly because liquid calories are poorly compensated?
Research question in detail (PICO)
- Population
- Children and adults
- Intervention or exposure
- SSB
- Comparison
- Non-caloric beverages or reduced SSB
- Geographic scope
- universal
- Outcomes
- Body weight/BMI
- Energy compensation
Study results as reported by the authors. Our assessment follows below.
What speaks for the claim?
Sugar drinks made children gain more weight
de Ruyter JC et al.. N Engl J Med 2012 · PMID 22998340 · DOI 10.1056/NEJMoa1203034 Randomised controlled trialabstract checkednewly addedDouble-blind 18-month RCT in 641 children: masked sugar-free drinks reduced weight gain by about 1 kg versus sugar drinks.
- Actual result
- BMI z diff CI -0.21 to -0.05; weight +6.35 vs +7.37 kg.
- Methodological quality
- high (blinded RCT)
- Relevance
- direct
- Limitations
- 26% discontinued.
Sugar and body weight: trials and cohorts
Meta-analysis of RCTs: adding sugars increases weight in adults eating freely.
- Actual result
- +0.75 kg (0.30-1.19).
- Methodological quality
- high
- Relevance
- direct
- Limitations
- Energy-mediated.
What speaks against the claim?
Cutting sugary drinks helped teenagers' weight for one year
Ebbeling CB et al.. N Engl J Med 2012 · PMID 22998339 · DOI 10.1056/NEJMoa1203388 Randomised controlled trialabstract checkednewly addedRCT in adolescents: reducing SSB lowered weight gain at 1 year but not at the 2-year primary endpoint.
- Actual result
- BMI -0.3 at 2 years (P=0.46).
- Methodological quality
- high
- Relevance
- direct
- Limitations
- Effect not sustained.
What does the overall evidence say?
Editorial assessment based on the studies above – not itself a study result.
Which position is better supported?
Sugary drinks cause extra weight gain; masked RCTs show incomplete compensation for liquid sugar.
Why some studies weigh more
Blinded RCT plus RCT meta-analyses outweigh mechanistic studies.
Is a causal relationship established? Yes, established Randomised evidence in children and adults.
Remaining uncertainty
Long-term persistence and size in adults.
Justified conclusion
Well supported.
Why do studies seem to disagree?
Ebbeling 2012 vs de Ruyter 2012: unblinded behavioural intervention with fading adherence versus masked beverage replacement; they answer different questions (programme effectiveness vs biological effect).
Does this depend on the country?
- General biological effects
- Universal.
A regional difference that changes the conclusion is not supported by robust data.
Statistical re-check
- Recalculated differences vs reported CIs – BMI z -0.13 and weight -1.02 kg match CI midpoints (-0.13; -1.01) recalculated
Certainty of the overall evidence (GRADE, simplified)
- Main outcome
- Body weight with SSB
- Starting level
- high (randomised trials)
- Reasons for downgrading
- indirectness (mechanism of satiety)
- Result
- moderate
Simplified assessment by a single reviewer, following the GRADE approach; not a formal GRADE panel. GRADE rates the body of evidence for one outcome, not individual studies.
Clinical guidelines and official statements checked
- WHO: Guideline: sugars intake for adults and children (2015), publication page and news release (I-SUG-05) (2015) · Original publication
Reduce free sugars to <10% of energy (strong); further reduction to <5% (~25 g) conditional. Free sugars include sugars added by manufacturer, cook or consumer and sugars naturally present in honey, syrups, fruit juices and concentrates (news release https://www.who.int/news/item/04-03-2015-who-calls-on-countries-to-reduce-sugars-intake-among-adults-and-children, fetched 2026-10-26). - ADA: 5. Facilitating Positive Health Behaviors and Well-being to Improve Health Outcomes: Standards of Care in Diabetes-2026 (Diabetes Care 2026;49 Suppl 1:S89-S131; I-INTC-03) (2026) · Original publication
Rec. 5.14: minimise sugar-sweetened beverages, sweets, refined grains, processed and ultra-processed foods. Rec. 5.21: encourage water over other beverages (A). Rec. 5.22: nonnutritive sweeteners can replace sugar-sweetened products if consumed in moderation and for the short term (B). Rec. 5.25: replace sugar-sweetened beverages (including any juices) with water or low/no-calorie beverages and minimise foods with added sugar (B).
Search strategy
Documented literature searches for this review (search strings in the original database syntax).
| Database | Date | Search | Looking for | Hits | Included |
|---|---|---|---|---|---|
| PubMed | 26 October 2026 | (de Ruyter JC[au] OR Ebbeling CB[au]) AND sugar-sweetened beverages AND N Engl J Med[ta] | both directions | 3 | |
| Europe PMC | 26 October 2026 | ("sugar-sweetened beverages") AND ("body weight" OR adiposity OR BMI) AND ("randomized controlled trials" OR "randomised controlled trials") AND ("meta-analysis") AND PUB_YEAR:[2019 TO 2026] | both directions | 1228 | |
| Europe PMC | 26 October 2026 | ("liquid calories" OR "beverage energy" OR "liquid versus solid" OR "liquid and solid") AND (compensation OR satiety OR "energy intake") AND ("meta-analysis" OR "systematic review") | both directions | 463 |
Open questions
None material.
How deeply was this checked?
9 linked studies: 1 checked in full text (0 of them with a structured risk-of-bias assessment), 8 on the basis of the abstract, 0 bibliographically only. No formal assessment with two independent reviewers and no independent human expert review has taken place yet.
Last scientific review: 26 October 2026
History of the verdict
In plain language
Drinking sugary soft drinks adds calories that people usually do not balance by eating less. Trials show that adding sugary drinks increases body weight and removing them helps prevent weight gain. The effect comes from the extra calories. Diet sodas do not contain these calories, so this argument does not apply to them.
Studies on this claim
Studies supporting the claim (8)
-
Sugary drinks and child obesity
Ludwig DS et al.. Lancet 2001 · PMID 11229668 · DOI 10.1016/S0140-6736(00)04041-1Cohort studycited by the Tayebat archiveabstract checkedEach extra daily sugary drink was linked with higher BMI and 60% higher odds of obesity.
Does the study support the specific claim? Yes · Methodological quality: moderate · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Is SSB consumption related to obesity in children?
- How was the study done?
Prospective observational study of schoolchildren over 19 months.
- How many people, animals or samples were studied?
548; Schoolchildren aged ~11.7 y
- What was actually found?
Per additional serving: BMI +0.24 kg/m2 (0.10-0.39); obesity OR 1.60 (1.14-2.24); baseline intake and BMI change +0.18 kg/m2 per daily serving (0.09-0.27).
- How large was the observed effect?
OR 1.60 per serving
- What do the results mean in everyday life?
Sugary drinks may contribute to child weight gain.
- What are the study's limitations?
Small sample; Short follow-up; Observational design; residual confounding possible; Self-reported dietary intake
- What does it say about Elawady's specific claim?
Supports C-SOD-02.
- Do the results apply to Germany and Europe?
Moderate.
- What can explicitly not be concluded from this study?
Causation.
-
Sugary drinks, weight and diabetes in 91,000 women
Schulze MB et al.. JAMA 2004 · PMID 15328324 · DOI 10.1001/jama.292.8.927Cohort studycited by the Tayebat archiveabstract checkedWomen drinking one or more sugary drinks daily had 83% higher diabetes risk; fruit punch similar.
Does the study support the specific claim? Yes · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Are SSB associated with weight change and T2D in women?
- How was the study done?
Prospective cohort (NHS II) 1991-1999.
- How many people, animals or samples were studied?
91,249 (diabetes); 51,603 (weight); 741 T2D cases; Women aged young/middle-aged
- What was actually found?
Increasing SSB to >=1/day: 4.69 and 4.20 kg weight gain over 4 years vs 1.34 and 0.15 kg in decreasers; >=1 SSB/day T2D RR 1.83 (1.42-2.36); fruit punch RR 2.00 (1.33-3.03).
- How large was the observed effect?
RR 1.83 (1.42-2.36)
- What do the results mean in everyday life?
Sugary drinks, including fruit punch, are linked with weight gain and diabetes.
- What are the study's limitations?
Observational design; residual confounding possible; Self-reported dietary intake
- What does it say about Elawady's specific claim?
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Causation.
-
Sugary drinks and weight gain: meta-analysis
Malik VS et al.. Am J Clin Nutr 2013 · PMID 23966427 · DOI 10.3945/ajcn.113.058362Meta-analysiscited by the Tayebat archiveabstract checkedAdding sugary drinks increased adult weight by about 0.85 kg in trials; cohorts show weight gain per serving.
Does the study support the specific claim? Yes · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Do SSBs promote weight gain?
- How was the study done?
Systematic review and meta-analysis of cohorts and RCTs.
- How many people, animals or samples were studied?
32 articles (cohorts n=25,745 children/174,252 adults; RCTs n=2,772/292); Children and adults
- What was actually found?
Cohorts: per daily serving BMI +0.06 (children) and +0.22 kg/year (adults, random effects); RCTs: children BMI -0.17 (-0.39 to 0.05) with reduction; adults +0.85 kg (0.50-1.20) with SSB addition.
- How large was the observed effect?
+0.85 kg (RCTs adults)
- What do the results mean in everyday life?
Sugary drinks add weight.
- What are the study's limitations?
Child RCT estimate imprecise (random effects CI crosses 0); Overlap with other reviews
- What does it say about Elawady's specific claim?
Supports C-SOD-02.
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Long-term size of effect.
-
Soda calories vs jelly-bean calories
DiMeglio DP, Mattes RD. Int J Obes Relat Metab Disord 2000 · PMID 10878689 · DOI 10.1038/sj.ijo.0801229Randomised controlled trialcited by the Tayebat archiveabstract checkedPeople ate less to balance jelly-bean calories but not soda calories, and gained weight only with soda.
Does the study support the specific claim? Yes · Methodological quality: low · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Do liquid and solid carbohydrate loads differ in dietary compensation?
- How was the study done?
Crossover feeding study (4-week periods).
- How many people, animals or samples were studied?
15; Healthy adults (7 men, 8 women)
- What was actually found?
Solid: compensation 118%; liquid: -17%; body weight and BMI increased only during liquid period.
- How large was the observed effect?
Compensation 118% vs -17%
- What do the results mean in everyday life?
Liquid sugar calories are less filling.
- What are the study's limitations?
Very small; Randomisation not described in abstract
- What does it say about Elawady's specific claim?
Supports C-SOD-02.
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Long-term effect.
-
Sweetened drinks dose-response
Qin P et al.. Eur J Epidemiol 2020 · PMID 32529512 · DOI 10.1007/s10654-020-00655-yMeta-analysiscited by the Tayebat archiveabstract checkedEach extra 250 mL/day of sugary drinks was linked with 12-19% higher risks of obesity and diabetes; diet drinks showed similar links.
Does the study support the specific claim? Yes · Methodological quality: moderate · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Dose-response of SSB/ASB with obesity, T2D, hypertension, mortality.
- How was the study done?
Dose-response meta-analysis of prospective cohorts.
- How many people, animals or samples were studied?
39 articles; Adults
- What was actually found?
Per 250 mL/day SSB vs ASB: obesity RR 1.12 vs 1.21; T2D 1.19 vs 1.15; hypertension 1.10 vs 1.08; all-cause mortality 1.04 vs 1.06 (with I2 up to 92.6%).
- How large was the observed effect?
T2D RR 1.19 per 250 mL/d (SSB)
- What do the results mean in everyday life?
More sweetened drinks, more risk in observational data.
- What are the study's limitations?
High heterogeneity; Cohorts only; Reverse causation for ASB
- What does it say about Elawady's specific claim?
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Causation, especially for diet drinks.
-
Sugar and body weight: trials and cohorts
Te Morenga L et al.. BMJ 2012 · PMID 23321486 · DOI 10.1136/bmj.e7492Meta-analysisindependent cross-checkabstract checked★ key studyEating more sugar increased body weight and eating less reduced it, through extra calories.
Does the study support the specific claim? No · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Is intake of dietary sugars associated with body weight?
- How was the study done?
Systematic review and meta-analysis of RCTs (>=2 weeks) and cohorts (>=1 year), Cochrane/GRADE methods (WHO-commissioned per literature; not stated in abstract).
- How many people, animals or samples were studied?
30 trials; 38 cohorts; Adults and children
- What was actually found?
Adults ad libitum: reduced sugars -0.80 kg (0.39-1.21); increased sugars +0.75 kg (0.30-1.19); isoenergetic exchange 0.04 kg (-0.04 to 0.13). Children: SSB highest vs lowest OR 1.55 (1.32-1.82) for overweight at 1 year.
- How large was the observed effect?
-0.80 kg with reduction; +0.75 kg with increase
- What do the results mean in everyday life?
More sugar means more weight for most people.
- What are the study's limitations?
Heterogeneity in one analysis; Child trials low compliance
- What does it say about Elawady's specific claim?
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
Effects in isoenergetic diets (none found).
-
WHO sugar guideline
World Health Organization. WHO news release 2015Guideline / position statementindependent cross-checkfull text checked★ key studyWHO recommends keeping free sugars, including honey and juice sugars, below 10% of energy.
Does the study support the specific claim? No · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Recommended limit for free sugars.
- How was the study done?
International guideline (news release read in full).
- How many people, animals or samples were studied?
Not applicable; Adults and children
- What was actually found?
Free sugars <10% of total energy (strong recommendation); further reduction <5% (~25 g, 6 teaspoons) conditional. Free sugars include sugars added by manufacturer, cook or consumer and sugars naturally present in honey, syrups, fruit juices and fruit juice concentrates; not sugars in fresh fruit, vegetables or milk. A can of sugar-sweetened soda contains up to 40 g free sugars.
- How large was the observed effect?
<10% energy (strong); <5% (conditional)
- What do the results mean in everyday life?
Honey, juices and table sugar all count.
- What are the study's limitations?
Conditional recommendation for <5%
- What does it say about Elawady's specific claim?
Defines honey and juices as free sugars: contradicts C-SUG-01 and C-SUG-06; supports C-SUG-02.
- Do the results apply to Germany and Europe?
Applies globally including EU and Egypt.
- What can explicitly not be concluded from this study?
Not applicable.
-
Sugar drinks made children gain more weight
de Ruyter JC et al.. N Engl J Med 2012 · PMID 22998340 · DOI 10.1056/NEJMoa1203034Randomised controlled trialindependent cross-checkabstract checked★ key studyIn an 18-month blinded trial, Dutch children given one sugary drink a day gained about 1 kg more than children given an identical-tasting sugar-free drink.
Sugar drinks caused ~1 kg more weight gain over 18 months versus masked sugar-free drinks.
Does the study support the specific claim? Yes · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Test whether replacing sugary drinks with sugar-free drinks reduces weight gain in children.
- How was the study done?
Children received one can a day of either drink for 18 months without knowing which.
- How many people, animals or samples were studied?
641 Dutch school children.
- What was actually found?
Less weight and fat gain with sugar-free drinks.
- How large was the observed effect?
Weight difference about -1.0 kg; BMI z -0.13.
- What do the results mean in everyday life?
Liquid sugar adds weight because children do not eat less to compensate.
- What are the study's limitations?
Many children stopped drinking; sensitivity analysis borderline.
- What does it say about Elawady's specific claim?
Supports the claim.
- Do the results apply to Germany and Europe?
Directly European.
- What can explicitly not be concluded from this study?
Cannot speak to adults or long-term disease.
No study in this group is marked as key.
Studies contradicting or limiting the claim (0)
No studies in this group.
Studies with mixed or unclear results (1)
-
Cutting sugary drinks helped teenagers' weight for one year
Ebbeling CB et al.. N Engl J Med 2012 · PMID 22998339 · DOI 10.1056/NEJMoa1203388Randomised controlled trialindependent cross-checkabstract checkedOverweight teenagers who replaced sugary drinks gained less weight during the one-year programme, but the difference was gone a year later.
Weight benefit at 1 year, not at 2 years.
Does the study support the specific claim? Partly · Methodological quality: high · Certainty of evidence: see overall assessment of the claim
Original publication · Scientific details
Study explained simply
- What did the researchers want to find out?
Test whether reducing sugary drinks slows weight gain in teenagers.
- How was the study done?
Randomised trial with free non-caloric drinks delivered home for a year.
- How many people, animals or samples were studied?
224 overweight teenagers.
- What was actually found?
Less weight gain at 1 year; no difference at 2 years.
- How large was the observed effect?
Weight -1.9 kg at 1 year.
- What do the results mean in everyday life?
Reducing sugary drinks helps while it is maintained.
- What are the study's limitations?
Effect did not last after the programme.
- What does it say about Elawady's specific claim?
Partly supports the claim.
- Do the results apply to Germany and Europe?
Moderate.
- What can explicitly not be concluded from this study?
Cannot show lasting benefit.
No study in this group is marked as key.
Regional context
Germany / EU
The WHO guideline (2015) on free sugars and EFSA's 2022 opinion (intake of added and free sugars 'as low as possible') apply in Germany and the EU; the European EPIC cohort (A-SOD-17) found associations in 10 European countries including Germany.
Egypt
Sugar-sweetened soft drinks are widely sold in Egypt. No Egypt-specific study was reviewed for this claim; the physiological mechanisms and cohort findings are not known to differ by country, but absolute risks depend on local consumption and disease rates.
What applies in my country?
General scientific assessment
Mostly true Sugary drinks promote weight gain through extra calories; this part of the archive's claim is well supported.
The following areas of food safety are relevant to this statement:
- Additives and processing
Honey, sugar, sweets
Choose a country to see the checked information for it.
Limits: legal rules describe what is allowed, not what every single food contains. Measurement data come from samples and are meaningful only together with their method details.
Country of residence or purchase Compare with another country
References
- Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis.. Ludwig DS et al.. Lancet 2001 · PMID 11229668 · DOI 10.1016/S0140-6736(00)04041-1
- Sugar-sweetened beverages, weight gain, and incidence of type 2 diabetes in young and middle-aged women.. Schulze MB et al.. JAMA 2004 · PMID 15328324 · DOI 10.1001/jama.292.8.927
- Sugar-sweetened beverages and weight gain in children and adults: a systematic review and meta-analysis.. Malik VS et al.. Am J Clin Nutr 2013 · PMID 23966427 · DOI 10.3945/ajcn.113.058362
- Liquid versus solid carbohydrate: effects on food intake and body weight.. DiMeglio DP, Mattes RD. Int J Obes Relat Metab Disord 2000 · PMID 10878689 · DOI 10.1038/sj.ijo.0801229
- Sugar and artificially sweetened beverages and risk of obesity, type 2 diabetes mellitus, hypertension, and all-cause mortality: a dose-response meta-analysis of prospective cohort studies.. Qin P et al.. Eur J Epidemiol 2020 · PMID 32529512 · DOI 10.1007/s10654-020-00655-y
- Dietary sugars and body weight: systematic review and meta-analyses of randomised controlled trials and cohort studies.. Te Morenga L et al.. BMJ 2012 · PMID 23321486 · DOI 10.1136/bmj.e7492
- WHO calls on countries to reduce sugars intake among adults and children (Guideline: Sugars intake for adults and children, 2015) — news release. World Health Organization. WHO news release 2015
- A trial of sugar-free or sugar-sweetened beverages and body weight in children.. de Ruyter JC et al.. N Engl J Med 2012 · PMID 22998340 · DOI 10.1056/NEJMoa1203034
- A randomized trial of sugar-sweetened beverages and adolescent body weight.. Ebbeling CB et al.. N Engl J Med 2012 · PMID 22998339 · DOI 10.1056/NEJMoa1203388