Replacing Sugary Drinks With Plain Coffee Or Tea May Reduce Diabetes Risk

Dr Nita Forouhi, MRCP, PhD, FFPHM MRC Programme Leader and Consultant Public Health Physician MRC Epidemiology Unit University of Cambridge School of Clinical Medicine Cambridge Biomedical Campus Cambridge Interview with:
Dr Nita Forouhi, MRCP, PhD, FFPHM
MRC Programme Leader and Consultant Public Health Physician
MRC Epidemiology Unit
University of Cambridge School of Clinical Medicine
Cambridge Biomedical Campus
Cambridge UK

Medical Research: What is the background for this study? What are the main findings?

Dr. Forouhi: Consumption of soft drinks is known to cause obesity and may contribute to the development of type 2 diabetes. We had previously published findings from the EPIC-InterAct study in 8 European countries that habitual consumption of sugar sweetened beverages increases the risk of developing type 2 diabetes, and we now wanted to probe deeper to understand more about this relationship between sweet beverages and diabetes.

We conducted research in the large EPIC-Norfolk study which included more than 25,000 men and women aged 40–79 years living in Norfolk, UK. Study participants recorded everything that they ate and drank for 7 consecutive days covering weekdays and weekend days, with particular attention to type, amount and frequency of consumption, and whether sugar was added by the participants. During approximately 11 years of follow-up, 847 study participants were diagnosed with new-onset type 2 diabetes.

By using this detailed information on diet, we were able to study several different types of sugary beverages, including sugar-sweetened soft drinks, sweetened tea or coffee and sweetened milk drinks as well as artificially sweetened beverages (ASB) – such as diet soft drinks – and fruit juice, and to examine what would happen if plain water, unsweetened tea or coffee or artificially sweetened beverages were substituted for sugary drinks.

Our study provided three main findings:

First, there was an approximately 22% increased relative risk of developing type 2 diabetes per extra serving per day habitually of each of soft drinks, sweetened milk beverages and ASB consumed, even after accounting for a range of important factors including other lifestyle and social factors and for total energy intake. However, after further accounting for body mass index and waist girth as markers of obesity, there remained a higher risk of diabetes associated with consumption of both soft drinks and sweetened milk drinks, but the link with ASB consumption no longer remained, possibly because artificially sweetened beverages was likely to be consumed by those who were already overweight or obese.

Second, when we estimated the likely effects of replacing a habitual serving of soft drinks with a serving of water or unsweetened tea or coffee, we found that the risk of diabetes could have been cut by 14%; and by replacing a habitual serving of sweetened milk beverage with water or unsweetened tea or coffee, that reduction could have been 20%–25%.  However, consuming ASB instead of any sugar-sweetened drink was not likely to reduce the risk of diabetes, when accounting for baseline obesity and total energy intake.

Third, we found that each 5% of higher intake of energy (as a proportion of total daily energy intake) from total sweet beverages (soft drinks, sweetened tea or coffee, sweetened milk beverages, fruit juice) was associated with a 18% higher risk of diabetes. We estimated that if study participants had reduced the energy they obtained from sweet beverages to below 10%, 5% or 2% of total daily energy, 3%, 7% or 15% respectively of new-onset type 2 diabetes cases could have been potentially avoided.

Medical Research: What should clinicians and patients take away from your report?

Dr. Forouhi: Our findings apply to people in the general population for the prevention of type 2 diabetes, and as such apply to adult consumers regardless of whether they are in a setting with their doctors or not. All healthcare and allied professional as well as public health bodies can provide advice about reducing the consumption of free sugars in the diet, in keeping with the most recent guidance from the World Health Organization. Limiting or eliminating the consumption of sugary beverages and replacing them with healthier alternatives is one of the easiest ways to achieve that goal.

In particular, replacing the habitual consumption of sugar sweetened beverages with healthier beverage options, like water or unsweetened tea or coffee, can play an important part in the prevention of type 2 diabetes.

This is the first report of an association between the habitual daily consumption of sweetened milk beverages and risk of new-onset type 2 diabetes. Past research typically considered sugar sweetened beverages as soft drinks such as fizzy drinks, colas, squashes and juice drinks, but our study raises the possibility that the habitual daily consumption of sweetened milk drinks such as flavoured milk and milkshakes may also elevate the risk of new-onset diabetes. Though this finding should be replicated in other studies, it certainly sounds a note of caution that the past focus solely on non-milk sweetened drinks may miss out an important class of sweetened beverages that could be related with adverse health effects through the non-milk free sugars.

Extra information: We acknowledge that studies such as ours that are observational in nature do have limitations including that they do not prove a cause and effect association, whilst randomised controlled trials, such as those used when evaluating new drugs, provide the best evidence of cause and effect. Realistically we are limited to observational studies as it is impractical to randomly recruit and retain people to follow particular diets or drinks for long periods of time until a health condition develops. So, observational studies such as the one we report now have an important role to play in providing a good evidence base. The current study has several strengths that contribute to its robustness of findings: it is based on a large sample size of people who did not initially have diabetes so we could capture new-onset cases, it assessed diet with the detailed 7-day food diary, and it accounted for several important factors such as lifestyles and social factors as well as energy intake and obesity, that could potentially influence the findings because they may be related to intake of sugary drinks and to diabetes risk. Thus overall, the findings from this study are robust, and contribute meaningfully to understanding the link between sugary beverages consumption and risk of type 2 diabetes.

Medical Research: What recommendations do you have for future research as a result of this study?

Dr. Forouhi: More research is needed to get a better understanding of the relationship between regularly consuming artificially sweetened beverages (ASB) such as diet drinks and risk of type 2 diabetes. This is a complex area to tease out because people who are already overweight or obese tend to consume ASB in preference to the sugary options as part of weight control, with potential for so-called reverse causation. In our study, once obesity was accounted for there was no demonstrable relationship between ASB consumption and type 2 diabetes, but replacing ASB for sugary drinks was not associated with a significant reduction in diabetes risk. This deserves further research to unravel the complicated relationships between these drinks, obesity, weight change and disease outcomes.

Extra information: The challenge is that observational studies are prone to alternative explanations for findings through confounding and bias, and though randomised clinical trials would provide better evidence, they are difficult and expensive to conduct for disease end points. Random recruitment and retention would be challenging as participants would need to consume specific foods/drinks and hold other factors “constant” over long periods of time till an endpoint develops; that would be more feasible for weight change than for a condition like type 2 diabetes or other “hard” endpoints. Trials of intermediate metabolic endpoints, such as glucose or insulin levels or lipid blood fats, that are risk markers of disease conditions can be an option. Further observational studies that can account for changes in diet and changes in weight over time could help too.

Further research should also address the relationship between sweetened milk drinks such as milkshakes, flavoured milks etc, and the risk of diabetes in different age groups, as our study was the first to report such an association in a long-term study of middle aged and older adults that followed them up over time for the development of diabetes.


O’Connor, L et al. Prospective associations and population impact of sweet beverage intake and type 2 diabetes, and effects of substitutions with alternative beverages. Diabetologia; 30 April 2015