People have been looking for a quick and easy way to lose weight for a good long while, probably for as long as an anorectic physique has been considered attractive. And what easier way could there be than taking a pill? No need to change your diet. No need to exercise (ineffective anyway, I know). Just pop that pill and be done with it!
The first truly effective weight loss pill was perhaps dinitrophenol, which came on the market in the early 1930’s. It had a pretty ingenious mechanism of action, which requires some explaining. In the normal state of things, most of the energy we consume through food is transferred to a molecule called ATP (adenosine triphosphate). ATP in turn functions kind of like a battery, powering most of the things that happen in our bodies.
Most ATP is produced in the mitochondria by a giant (on the molecular scale) enzyme called ATP synthase, that sits on the mitochondrial membrane and looks (and functions) a bit like a wind turbine. ATP synthase is driven by an osmotic gradient, caused by a difference in the number of protons on each side of the mitochondrial membrane. As protons move along the gradient, they cross the mitochondrial membrane, passing through ATP synthase, and ATP is generated.
Dinitrophenol allows protons to bypass ATP synthase, and move across the membrane without transferring energy to ATP. Since energy can’t be destroyed, it is instead released as heat. So the drug basically works by making your metabolism less efficient. More of the energy that you consume ends up being given off as heat rather than being used to power reactions in your body. It’s a bit like replacing a modern low energy light bulb with the old incandescent type.
For a few years, dinitrophenol was very popular. But it ran in to one big problem. It was a bit too easy for people to overdose on it. Occasionally people would take too much of the drug, causing their body temperatures to spike to dangerously high levels, and this would cause them to spontaneously combust. Ok, that’s a bit of an exaggeration, but they died. So the drug stopped being used in the late 1930’s.
After dinitrophenol fell out of use, amphetamines became popular as a weight loss drug for a couple of decades. Among the many effects amphetamines have on the nervous system, they suppress appetite, which makes them pretty effective at helping people to lose weight. But their addictiveness eventually caused them to be classed as narcotics, and so they were also largely taken off the market. People who wanted to lose weight by popping a pill were back at square one.
At present, there is only one drug approved for use as a treatment of obesity, and that is orlistat, a drug that came on the market around the turn of the new millennium. Orlistat inhibits pancreatic enzymes involved in breaking down fats to a form that allows them to be absorbed in to the body. This causes the fats to stay in the intestine, and then pass out with the stool.
Unfortunately, orlistat is pretty useless, at best causing 2 to 3 kilograms of weight loss after a full year of use. And the side effects, which consist of fatty, oily stools and increased farting, aren’t nice. Additionally, people taking orlistat frequently become deficient in the fat soluble vitamins (A, D, E, and K).
Now, amphetamines aren’t the only drugs that over the years have been noticed to have weight loss as a common side effect. Levothyroxine, used to treat hypothyroidism, induces weight loss by speeding up metabolism. Unfortunately the risks associated with its use mean that it can’t be used as a weight loss pill.
Then there’s the GLP-1 analogues. GLP-1 is a hormone that is released when we eat food. It has been found to increase insulin levels and decrease glucagon levels, both of which cause blood sugar to drop. For this reason, GLP-1 was at one time heavily researched as a treatment for type 2 diabetes, and a number of GLP-1 analogues (synthetic drugs that have a physiological effect similar to GLP-1) were developed and marketed for this purpose.
The fist GLP-1 analogue to come on the market was exenatide, which was approved for use in type 2 diabetes in 2005. Unfortunately for pharmaceutical companies, it still isn’t clear whether these drugs actually make people with type 2 diabetes live longer, so they have remained a second-line agent, after older (and cheaper) generic drugs like metformin and sulfonylurea. Fortunately for pharmaceutical companies, the GLP-1 analogues were found to have one beneficial side effect that isn’t seen with either metformin or sulfonylurea – they cause people to lose weight.
Naturally, this has led to the suggestion that they could be used as weight loss drugs in people who are obese but who don’t have diabetes. Which leads us to a study that was recently published in the New England Journal of Medicine.
This was a double-blind randomized controlled trial carried out at over a hundred different sites in 16 different countries. 1,961 participants were recruited and randomized to either a GLP-1 analogue called semaglutide or placebo. The study was funded by Novo Nordisk, which holds the patent on semaglutide.
In order to be included in the study, participants had to be over 18 years old and they had to either have a BMI (Body Mass Index) over 30 (the cut-off for obesity), or over 27 and also have some weight related co-morbidity (e.g. high blood pressure, high cholesterol levels, sleep apnea, or heart disease). People with diabetes were excluded, since earlier studies have already shown that GLP-1 analogues are effective at reducing weight in people with type 2 diabetes. People who had previously suffered from pancreatitis (inflammation of the pancreas) were also excluded, since GLP-1 analogues have previously been shown to increase the risk of pancreatitis. The study also excluded people who had recently been taking some other form of weight loss drug, or who had had weight loss surgery.
Participants received a once-weekly injection of either semaglutide or placebo. All participants also received traditional weight loss councelling once every four weeks, in which they were recommended to follow a low calorie/low fat diet and take regular exercise. They were followed for 68 weeks. The reason 68 weeks was chosen as the length of the study was because there was a 16 week run-in period, during which the dose of semaglutide was gradually titrated up, followed by 52 weeks with the full dose (or a lower dose if a participant was not able to tolerate the full dose).
What can we say so far? This was a well done, high quality study, with lots of participants, which provides a high level of statistical power. Usually studies aim for a power of 80%, which means they have an 80% chance of finding a real difference if one exists. This study had a power of 99%! They recruited participants at lots of different sites in several different countries, which minimizes the risk that some condition specific to one setting will confound the results. The intervention received by the two groups was identical, with the one exception of getting either semaglutide or a placebo. And the participants were followed for over a year.
The average age of the participants was 46 years. The average BMI at the start of the trial was 38, and the average weight was 105 kilograms (231 pounds).
Ok, let’s get to the results. And let’s start with the good news.
At the end of the study, participants in the semaglutide arm had lost 14,9% of their body weight. In the placebo arm, participants had lost 2,4% of their body weight. That means an average weight loss of 15,3 kilograms (34 pounds) in the treatment group, and 2,6 kilograms (5,7 pounds) in the placebo group. The absolute difference in weight loss between the two groups was a 12,4%. That is a big, impressive difference. On average, participants in the treatment arm lost 13 kilograms (28 pounds) more than people in the placebo arm.
Overall, 69% of participants in the semaglutide group lost at least 10% of their body weight, compared with only 12% of participants in the placebo group. Participants in the semaglutide group also saw a much bigger reduction in their waist circumference, losing 13,5 cm, as compared with only 4,1 cm in the placebo group.
Now let’s get to the bad news.
74% of participants in the study were female, and 75% were white. That means the overall results mainly apply to white females. Unfortunately, Novo Nordisk don’t provide a breakdown of the results by either gender or by ethnicity, so it’s impossible to tell whether the benefits seen here apply equally across genders and ethnicities, or whether the benefit is only seen in white females. It’s odd that they don’t provide this information, and the suspicious little devil sitting on my shoulder is whispering in to my ear that that probably means semaglutide only works for white women.
And now, let’s get to the really bad news.
9,8% of participants in the semaglutide arm suffered a serious adverse event, as compared with 6,4% in the placebo gorup. That gives an absolute difference of 3,4%, which would mean that roughly one in thirty people taking the drug for a year can expect to suffer a serious adverse event as a result of taking it. Not good. So, what were these serious adverse events?
Unfortunately Novo Nordisk again aren’t kind enough to provide us with a clear breakdown of what the serious adverse events were, although they say that they mainly consisted of gastrointestinal disorders and hepatobiliary disorders (disorders to do with the liver and gall bladder).
In order to be classified as “serious”, an adverse event generally has to either result in death or at least be potentially life-threatening, or result in hospitalization, or result in permanent disability. In the study, one person died in the treatment group and one person died in the placebo group, so we know that at least the semaglutide wasn’t killing people. Most likely the drug was resulting in an increase in hospitalizations, for example due to gall stones (a hepatobiliary disorder) or dehydration due to vomiting (a gastrointestinal disorder).
It’s unfortunate that Novo Nordisk don’t provide more detail on what the serious adverse events were when there is such a clear signal of harm. Nor do they tell us how many people were hospitalized over the course of the study in each treatment arm. I listened to a half hour interview with one of the lead researchers in charge of this study, and funnily enough he never once mentioned that there was a significant increase in serious adverse events in the treatment group. In fact, he said explicitly that the drug was largely side-effect free, with the exception of some mild nausea that usually resolved within a few weeks of beginning treatment. For those who like relative risk numbers (pharmaceutical companies certainly do if they make their drugs look good), the relative risk of a serious adverse event increased by 53% in the group receiving semaglutide!
So we can conclude that treatment with semaglutide results in some type of gastrointestinal or hepatobiliary illness that is serious enough to at the very least require a trip to the hospital in one person in thirty per year of treatment. If we look at adverse events more generally (not just serious ones), we see that treatment with semaglutide resulted in a significant increase in nausea (44% vs 17%), diarrhea (32% vs 16%), vomiting (25% vs 7%), and constipation (23% vs 10%).
Hmm, suddenly semaglutide isn’t looking quite so good. 7% of participants in the semaglutide group chose to discontinue treatment due to adverse events, as compared with 3% in the placebo group.
Ok, so what can we conclude from all this?
Firstly, semaglutide does appear to be highly effective at helping people to lose weight (at least if those people are white women). As mentioned earlier, the only currently approved weight loss drug on the market is orlistat, which results in 2-3 kilograms of weight loss over the couse of a year. Semaglutide appears to blow orlistat out of the water, resulting in four to five times the amount of weight loss.
Secondly, this study confirms yet again how utterly useless traditional advice, to follow a low fat/low calorie diet and exercise more, is at helping people to lose weight. The placebo group lost less than three kilograms, in spite of being motivated enough to take part in a study and receiving coaching on a regular basis.
Thirdly, semaglutide has a significant side effect profile, resulting in a one in thirty chance of suffering a serious adverse event over the course of a year of treatment, and causing milder gastrointestinal side effects, like nausea, vomiting, and diarrhea, in a large proportion of those taking the drug.
Personally, I’m not a fan of medicalizing lifestyle issues, and using drugs to accomplish what could be accomplished with a change of diet. As I wrote about recently, a ketogenic diet is effective at inducing weight loss (far more so than a low fat/low calorie diet), and it is safe, and should in my opinion be the first line intervention for people who want to lose weight.
For those who don’t achieve sufficient weight loss with a ketogenic diet, or who aren’t willing to try it for whatever reason, semaglutide (or another GLP-1 analogue) could be an option, certainly before taking the more drastic step of doing bariatric surgery. However, patients should be informed about the significant side effect profile of the drug.