Do diet soft drinks actually make you gain weight?

NHS: "Is Diet Coke making you fat? People who drink at least one can a day have larger waist measurements," the Mail Online reports. A US study found an association between the daily consumption of diet fizzy drinks and expanded waist size. This study included a group of older adults aged 65 or over from San Antonio, Texas. Researchers asked participants about their consumption of diet soft drinks and measured their body mass index (BMI) and waist circumference. They then looked at whether this was associated with changes in body measures over the next nine years.
The study found people who drank diet soft drinks every day had a greater increase in waist circumference at later assessments compared with those who never drank them (3.04cm gain versus 0.77cm). Daily drinkers also had a slight gain in BMI (+0.05kg/m2) compared with a minimal loss in non-drinkers (-0.41kg/m2).

The hypothesis that diet drinks can actually make you fatter is not a new one – we covered a similar study back in January 2014. The problem with this field of research is it is very difficult to prove cause and effect. As with this study, people who regularly drink diet drinks may be overweight to start with and they could be drinking diet drinks in an effort to lose weight.
This study will add to the variety of research examining the potential harms or benefits of artificial sweeteners or diet drinks. But it does not prove that drinking diet drinks will make you fat.
If you are trying to lose weight, good old-fashioned tap water is a cheaper, calorie-free alternative to diet drinks.

Where did the story come from?

The study was carried out by researchers from the University of Texas Health Science Center in the US, and was funded by the US National Institute on Aging, the National Institute of Diabetes and Digestive and Kidney Diseases, and the National Center for Research Resources. The authors declare no conflicts of interest. 
It was published in the peer-reviewed Journal of the American Geriatrics Society.
The Mail Online's coverage of this study seems overly conclusive, suggesting it provides evidence that drinking diet fizzy drinks causes people to become overweight. But this has not been proven, and the Mail did not consider this study's many limitations in their reporting.
It also includes an error in its story, describing the study of 749 people "in which 466 participants survived". This is the number of people who had data on body measurements available for at least one of the follow-up assessments. It is the retention of people in the study, not the survival rate.
Furthermore, in saying that, "Large waistlines linked to diabetes, stroke, heart attack and cancer", it is suggested that this study found a higher waist circumference was linked to the development of these diseases. However, health outcomes were not assessed in this study.
And, somewhat unfairly, Diet Coke was singled out as the main culprit. The study actually included any kind and brand of diet fizzy drink.

What kind of research was this?

This was a prospective cohort study that aimed to look at the link between diet soft drink intake and waist circumference.
The researchers discuss how concerns about high sugar intake over the past few decades have led to an increase in the consumption of artificial sweeteners. But the potential detrimental health effects of sweeteners have often been debated.
Some studies found no evidence for either the benefits or harms of sweeteners and diet drinks, while others found an increased risk of cardiovascular and metabolic risk factors, such as causing weight gain, leading to obesity, high blood pressure and diabetes.
This study aimed to examine the effect artificially sweetened diet drinks have on weight changes over time by looking at people taking part in an ongoing cohort study.
The main limitation with this type of study, however, is that it is not able to prove cause and effect, as the relationship is likely to be influenced by various other factors (confounders).

What did the research involve?

This research included a group of older Mexican and European American people taking part in the San Antonio Longitudinal Study of Aging (SALSA). This community-based study aimed to look at cardiovascular risk factors in people who were aged 65 or over at the start of the study (1992-96).
The first follow-up assessments were conducted an average of seven years later (2000-01), with two further follow-ups at 1.5-year intervals (2001-03, then 2003-04). The study included 749 people, with an average follow-up time of 9.4 years.
Assessments included measurements of participants' height, weight, waist circumference, fasting blood glucose levels, physical activity, and presence of diabetes. Dietary questionnaires were given at baseline and included the consumption of diet soft drinks.
People were asked the number of cans or bottles of diet soft drinks they consumed a day, week, month or year, and were categorised into three intake groups: non-users, occasional users (more than zero but less than one a day), and daily users (more than one a day) of diet soft drinks.
The researchers looked at the relationship between diet fizzy drink intake at the start of the study, and changes in BMI and waist circumference from when the study started to each follow-up point. Analyses were adjusted for age, gender, ethnicity, socio-demographics, diabetes, smoking status, and leisure activity.
Despite the large initial cohort size, only 384 people (51%) had data available on soft drink intake at baseline and body measurements at the first and second follow-ups, reducing to 291 (39%) by the third follow-up. 

What were the basic results?

The researchers found people who drank diet drinks at the start of the study also had significantly higher BMIs at the beginning of the study compared with non-users. They also tended to have higher waist circumference compared with non-users, though not significantly so.
The proportion of daily users who were overweight or obese at the start of the study was 88%, compared with 81% of occasional users and 72% of non-users.
Overall, the researchers found that for people who returned for one or more follow-ups, changes in BMI varied according to diet soft drink intake. Non-users experienced a minimal decrease in BMI (average 0.41kg/m2 decrease), as did occasional users (0.11kg/m2 decrease), while daily users had a slight increase (0.05kg/m2 gain).
Changes in waist circumference, meanwhile, were much more notable, with daily diet soft drink users experiencing a gain four times that of non-users. Average waist circumference gains at each interval were 0.77cm for non-users, 1.76cm for occasional users, and 3.04cm for daily users. 

How did the researchers interpret the results?

The researchers concluded that, "In a striking dose-response relationship, increasing diet soda intake was associated with escalating abdominal obesity, a potential pathway for cardiometabolic risk in this ageing population."

Conclusion

This prospective study found that people who drank diet soft drinks every day experienced greater waist circumference gain over up to nine years of follow-up compared with those who never drank diet drinks (3.04cm gain versus 0.77cm).
They also experienced a minimal gain in BMI (+0.05kg/m2) over follow-up, compared with a minimal loss in non-users of diet drinks (-0.41kg/m2).
However, this study certainly does not prove that diet drinks, and diet drinks alone, are responsible for these small increases in waist circumference and BMI.
People who drank diet drinks tended to have higher BMIs and waist circumferences than non-users to start with. At the start of the study, when diet soft drink consumption was assessed, 88% of those drinking them daily were overweight or obese, compared with 72% who weren't drinking soft drinks.
Though these people experienced slightly greater gains in BMI and waist circumference, these people tended to have generally higher body measurements to start with. It is possible that people with weight concerns may consume diet drinks in an effort to try to manage their weight.
There may be a variety of unhealthy lifestyle behaviours that contributed to the gain in body measures during the study. For example, the researchers adjusted their analyses for leisure-time physical activity, but did not consider food intake, apart from diet drinks, or look at total energy intake.
Overall, it is not possible to say from this analysis that the diet drinks are the cause of the changes in body measures, as various other unmeasured health and lifestyle factors could be having an influence.
Other points to bear in mind with this study are:

• This was an older age cohort of people above 65, so we don't know how representative the results would be for younger groups.
• This was a specific sample of people from San Antonio in Texas, and we don't know whether their health, lifestyle and environmental influences may differ from other population groups.
• Despite the initial sample size being fairly large at 749, data on drink consumption and body measurements was only available for about half of these people. The results may have been different had data been available for the full cohort.
• We don't know the significance of the small changes in BMI and waist circumference observed.
• We don't know whether continued daily consumption of diet soft drinks in the longer term would be associated with continuously increasing body measures, or whether this would have direct health effects (such as in terms of cardiovascular disease).
• The effects observed in this study can't be attributed to specific artificial sweeteners or specific diet soft drink brands.

The researchers' statement that there is a "striking dose-response relationship" between soda consumption and obesity seems overly bold given this study's limitations.
This study does not prove that drinking diet drinks will cause you to become fat. If you are trying to lose weight, we recommend that you ditch the expensive diet drinks and stick to water.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter. Join the Healthy Evidence forum.