 |
| Eight obese women (each have a BMI of 30) with different distributions of that weight and different waist sizes. Image from Wikimedia Commons. |
Today's article, published BBC, is about a paper just published in The Lancet Diabetes & Endocrinology that investigates the incidence rate of type 2 diabetes (T2D) in individuals who have undergone bariatric (weight loss) surgery. The news article is very brief, and states simply that researchers found an 80% reduction in the risk of T2D in individuals who have had weight loss surgery. The article also tells us that the UK NHS (their universal health insurer) is considering expanding the availability of covered weight loss surgery based on these findings. Let's see what the paper's authors actually say; does this new research show that bariatric surgery is a viable intervention to prevent or cure T2D?
Last week I used some pixels to talk about what T2D was, particularly with respect to how it is distinct from type 1 diabetes. It is well established in the medical community and the population at large that there is a strong link between obesity (a BMI over 30; BMI = (your weight in kilograms)/(your height in meters)2) and risk of developing T2D. Within the research community there is still a fair bit of disagreement as to the nature of that link. The conventional wisdom is that obesity causes T2D, but some research has suggested that sub-clinical T2D causes obesity (i.e. weight gain is the first symptom of T2D). Based on the conventional wisdom, many people with T2D are told to loose weight to control and potentially reverse their T2D. The countervailing theory leads to the idea that interventions traditionally used for weight loss, such as reduced calorie intake (particularly calories from sugars) and increased exercise, help the miss-regulated metabolism use glucose more appropriately and improve control of T2D irrespective of any change in body weight. This theory also posits that lifestyle changes must be permanent because we have not fixed the underlying miss-regulation, and with the incredibly high diet failure rate this makes the advice to patients basically the same (get on a diet and exercise program and stay there) regardless of the true connection between obesity and T2D. A permanent diet is a bitter pill for most patients to swallow, and compliance is a real problem. Because of this, researchers started investigating the potential of bariatric surgery as a possible treatment for T2D. I will state my bias regarding bariatric surgery here before getting into the data presented in today's journal article. I am very suspicious of any attempts to expand the definition of "people who could benefit from the procedure" when it comes to gastric bypass surgeries. Gastric bypass makes major irreversible changes to the way a persons body absorbs and metabolizes food. Individuals who've had the procedure often must take large doses of vitamins because they can no longer absorb nutrition from food at normal efficiency, the procedure has a high rate of complications (up to 3% mortality and 10% other serious complications, depending on procedure type), and it doesn't even work all that well. A 2007 study of nearly 800 patients found that 50% had regained most or all of the lost weight by 2 years post-procedure. I will say, 50% failure after 2 years is better than the failure rate for traditional "diet and exercise", but I feel the potential for serious complications makes these procedures only appropriate for a very small subset of obese individuals. However, as a scientist it is a matter of personal honor and integrity that I am willing to look at all new data objectively with an open (and critical) mind, and be willing to change my point of view when the data is convincing.
So what did the authors of this paper in The Lancet Diabetes & Endocrinology find out about T2D and bariatric surgery?
Today's research was done by researchers in the UK. They collected medical records from the Clinical Practice Research Datalink database, which includes 5 million patients in the UK who have agreed to open their medical records to researchers. From this database they pulled about 2100 individuals (84% women) who had a bariatric surgery (laparoscopic gastric banding, gastric bypass, or sleeve gastrectomy) between 2002 and 2014 and did not have symptoms of T2D at the time of surgery. This treatment group was then matched to a control group of obese individuals who were matched for age, weight, sex, and lack of T2D diagnosis, but never had bariatric surgery. They then followed the medical records of these two groups for 7 years post surgery/addition to study or until late 2014 (participants were followed an average of 2.8 years). They looked for diagnoses indicating patients developed T2D. They included any prescription for a glucose control drug (such as metaformin) unless it was accompanied by a PCOS diagnosis, and they noted if the diabetes was or was not during pregnancy. Despite the researchers efforts to create a match between the treatment and control groups, several statistically significant differences between the two groups were identified. The treatment group was statistically significantly more likely to have been diagnosed with depression, high blood pressure, and high cholesterol. This may be the result of only "sicker" obese individuals being recommended for bariatric surgery and it is controlled for in the data analyses.
By the end of the study 4.3% of the treatment group had developed T2D compared to 16.2% of the control group. When the variable following times were taken in account that represents an incidence rate of 5.7 diagnoses per 1000 persons per year for the treatment group and 28.2 diagnoses per 1000 persons per year for the control group. This means the hazard ratio for T2D diagnosis is 0.20 (95% confidence interval of 0.14-0.30), or that treated individuals developed T2D at 20% the rate of control individuals. It is important to note that the control group received no interventions to prevent T2D, they were not counseled regarding diet or exercise. In fact the researchers note that the control group has more missing data values for health markers such as blood pressure or cholesterol, suggesting that they were going to the doctor less often than the treatment group. They note these missing values could bias the study. I will add that the study did not (could not) look at lifestyle choices. I imagine that before being offered bariatric surgery the treatment group must have shown good compliance to lifestyle "best practices" such as good diet and exercise without improvement in health markers; they likely continued to comply to doctors' recommendations post-op. The control group would likely be a mixture of "good" patients and "naughty" patients, who do not eat well and get moderate exercise. This would introduce sever bias. However, based on the strength of the statistics the researchers found, they conclude that, "our findings, together with those of previous studies, suggest that bariatric surgery could be a highly effective method for prevention of diabetes in patients with severe obesity.... [However,] Further research is needed to understand the outcomes of different levels of uptake of obesity surgery, and the long-term effects for patients who receive current surgical procedures for obesity."
What are my conclusions?
I think it is important to do some more follow up studies that specifically control for adherence to lifestyle best practices for both treatment (bariatric surgery) and control (no surgery) groups. I would like to see a study that answers the question, is bariatric surgery better than early lifestyle intervention for the prevention of T2D? I would like the see that answered with good and poor lifestyle habits in bariatric surgery group compared separately, this would help us understand if people who can not or will not comply with best practices can be helped by bariatric surgery, or if surgery only helps those who comply with best practices. Another important question is if all types of surgery work equally well, because the complications for different types are very different. I think these are important questions to answer before changing which individuals are recommended for bariatric surgery.
I am not alone in my concerns regarding the findings of this paper. The very same issue of The Lancet Diabetes & Endocrinology has a comment article (think Letter to the Editor) regarding some of the weaknesses of the study. The comment article author, Dr. Himpens, shares my concern regarding the medical care of the control group. He also expresses concern about the short average follow-up (2.8 years) given the fact that many bariatric surgery patients start regaining weight around 24 months post-surgery. A separate study looking at bariatric surgery and T2D followed patients for 8 years; this study found that bariatric surgery initially resulted in weight loss and T2D remission, but 40% of patients regained the weight and T2D diagnosis by the end of the study. Additionally the study did not track the weight of all participants; in spite of the current knowledge that weight is correlated with T2D risk. He feels this is particularly important because the thinnest patients (BMI 30-34.9) show a non-statstically significant reduction in T2D risk, these patients are also the ones likely to loose the least weight as the result of bariatric surgery (they have the least weight to loose, maybe only 40lbs for a 5'8" person). These weaknesses in the study mean that much more research will be needed to discover if bariatric surgery can help with T2D separately from changes in lifestyle or body wight.
For me, based on the current state of research, I would not recommend one of these procedures for the sole purpose of preventing T2D, particularly in overweight (BMI 25-29.9) or mildly obese (BMI 30-34.5) individuals. But would encourage people worried about a family history of T2D to focus on taking good care of their bodies through exercise and a diet low in refined carbohydrate. Of course, I'm not that kind of doctor, and such matters should be a discussion between a patient and his/her medical doctor. Not a scientist on the internet. With respect to the BBC article, they really didn't say much about the study. So it's hard to make a statement regarding the quality of the science journalism beyond a D- for effort. I do feel that the omission of all the details of the study makes the findings sound more robust than they really are. The details are always important in science.
References (gray citations are behind a pay wall)
Booth, Helen, et. al. "Incidence of type 2 diabetes after bariatric surgery: population-based matched cohort study." The Lancet Diabetes & Endocrinology Online Nov 3, 2014.
Himpens, Jacques. "Can we safely state that bariatric surgery helps prevent type 2 diabetes?" The Lancet Diabetes & Endocrinology Online Nov 3, 2014.
Wang, JiaWei.
"Consumption of added sugars and development of metabolic syndrome
components among a sample of youth at risk of obesity." Applied Physiology, Nutrition, and Metabolism 39.4 (2014): 512-512.
Freeman, Joel B., and Heather Burchett. "Failure rate with gastric partitioning for morbid obesity." The American Journal of Surgery 145.1 (1983): 113-119.
Magro, Daniéla Oliveira, et al. "Long-term weight regain after gastric bypass: a 5-year prospective study." Obesity Surgery 18.6 (2008): 648-651.
No comments:
Post a Comment