Disclaimer: these are my own opinions, personal interpretations and are for information purposes only, reader discretion advised.
On the 8th of February NICE (National Institute of Clinical Evidence) approved the use of semaglutide as a “weight loss” intervention.
If you’ve been following me for a while, you’ll already know some of my likely thoughts on this.
But, I wanted to look into the evidence behind this.
First of all, what is semaglutide?
This is an injection that has been used for diabetes, it essentially helps a person secrete more insulin. It is a glucagon-like peptide-1 analogue. Simply, insulin is the hormone to lower blood sugar and insulin resistance or lack of insulin is the driving factor in diabetes.
What’s all the hype about?
A clinical trial published in the New England Journal of medicine; the Obesity STEP 1 Study, https://www.nejm.org/doi/full/10.1056/NEJMoa2032183
Once-weekly Semaglutide in Adults with “Overweight or Obesity” concluded semaglutide once weekly, plus lifestyle changes was associated with sustained weight loss.
What is NICE? this is an independent organisation that decides which drugs/ treatments are to be made available on the NHS. The aim is to make care more equitable, provide evidence based care that is cost effective and safe. I’ll return to this shortly.
Back to the evidence:
The STEP 1 study aimed to investigate the effectiveness of semaglutide on weight loss in people who are defined by the BMI criteria as overweight.
Participants were eligible for the trial if:
- They were 18 or over (pretty standard for most medical studies)
- **with self reported unsuccessful attempts at weight loss efforts
- A BMI of 30 or above
- A BMI of 27 or above with “an existing weight related health problem” these included high blood pressure, high cholesterol, heart disease or sleep apnoea
- BMI >40 with Dexa scan prior
Participants were excluded (not eligible to participate in the trial) if:
- they had a history of diabetes or pre diabetes
- Had taken glucose (sugar) lowering medications 3 months before trial start
- Treatment with similar medication to semaglutide 6 months prior
- Had a history of depression (2 years) prior
- Major psychiatric conditions defined as: schizophrenia or bipolar disorder
- Previous suicidal ideation (1 month before trial), attempt (lifetime)
- Alternative medicine supplement use 3 months prior
- Elevated markers of inflammation measured by raised calcitonin
- Genetic disorders in first degree relatives
- Impaired renal (kidney) function
- History of cancer (5 years)
- Strokes, TIA, myocardial infarction (heart attack) 1 month prior
- Pregnant, breast feeding or intending pregnancy
- Investigators’ opinion where something not covered by exclusions might interfere with participants safety or participation in the trial
So, what was the obesity START-1 trial?
A phase 3 randomised double blind, placebo-controlled trial. (what is that in English!?)
Basically before an intervention or drug can be adopted it undergoes rigorous testing, the gold standard of evidence in medicine comes from good quality trials. Phases of trials indicate both information about the size and stage of the evidence.
For most medications that make it into practice they’ve been through phases, at least phase 2 and ideally phase 3, which investigate their effect (phase 1 generally is small numbers looking at safety, phase 2 dosage before finally phase 3 the effectiveness compared to xyz). Double blind Randomised control trials mean neither the investigators or the participants know whether they are receiving the intervention (drug and In this case semaglutide) or the placebo (dummy drug). By conducting trials this way it avoids biased results (in theory).
This trial took place in 16 countries, 129 sites (a broad sample in theory). 1961 participants were included in the trial (again not small numbers)
Once enrolled participants were randomised (selected at random by a special software) to receive either semaglutide or placebo. 1306 participants were randomised to semaglutide vs 655 randomised to receive the placebo (this is a common method of trials where people are randomly allocated 2:1 fashion in the drug arm and placebo) although most trials participants are allocated 1:1.
They gave the participants the semaglutide or placebo weekly for 68 weeks as well as lifestyle interventions. After the 68 weeks participants had a 7 week period of no semaglutide, placebo or lifestyle intervention.
Lifestyle changes prescribed to participants included 500-calorie daily deficit relative to activity, 150 minutes of physical activity a week and recording daily of both activity and intake.
The main outcome measurements were achievement of 5% or more reduction in body weight from baseline and percentage difference from baseline in body weight at week 68. Other secondary outcomes ( measurements that were not the central focus of the study) included: achievement of reduction of more than 10%, 15%of body weight from baseline, systolic blood pressure, cholesterol levels, HBA1c various functional questionnaire’s Including physical activity, quality of life.
What were the results?
At the 68 week weight check there was a -14.9% change in body weight in the semaglutide group compared to -2.4% with the placebo. This was statistically significant ( the chance that this would be demonstrated again under the same circumstances and not as a result of chance). There were more reductions in weight in participants who received semaglutide than the placebo. The percentage reductions of 5%, 10, & 15% from baseline weight were all higher in the semaglutide group. However, although included in the results, reductions in cardiovascular metabolic risk factors such as HBA1c, cholesterol levels were not statistically different between those receiving semaglutide compared with placebo. Additionally more participants receiving semaglutide discontinued the treatment owing to adverse events (side effects) from the medication. The most common being gastrointestinal side effects including, nausea and diarrhoea.
What were the results?
At the 68 week weight check there was a -14.9% change in body weight in the semaglutide group compared to -2.4% with the placebo. This was statistically significant ( the chance that this would be demonstrated again under the same circumstances and not as a result of chance). There were more reductions in weight in participants who received semaglutide than the placebo. The percentage reductions of 5%, 10, & 15% from baseline weight were all higher in the semaglutide group. However, although included in the results reductions in cardiovascular metabolic risk factors such as HBA1c, cholesterol levels were not statistically different between those receiving semaglutide compared with placebo. Additionally more participants receiving semaglutide discontinued the treatment owing to adverse events (side effects) from the medication. The most common being gastrointestinal side effects including, nausea and diarrhoea.
Here’s where I think it’s important to know how to understand where evidence from trials can be misleading. On the face of it, this looks like a wonder drug for obesity. ( I say this with a bitter taste in my mouth because that in itself is a stigmatising sentence). The whole pretence of this study is filled with weight stigma, it is medicalising fatness. From the inclusion criteria you can see many of the participants were included by being fat, but without any “co-existing fat associated” conditions. They were included simply based upon their BMI, which is an unhelpful assessment of health. If I believed what the media reports about this and I hadn’t have been fortunate enough to a understand that fat does not equate to unhealthy and second my medical training that enables me to appraise the findings of such research I would have been led to believe two things: all fat much be medically treated and this is an effective method in doing so.
And so lets look closer at the trial design itself:
Most trials that are considered refutable minimise biases (any factor that may make the results difficult to interpret or apply to a wide population).
Let’s delve a bit deeper into the participants included in the trial:
Both groups of participants were fairly well matched with respect to age, BMI, pre-existing problems, diabetic status, ethnicity and gender. This sounds good at this point. But on a deeper look the participants in the trial were predominantly white female 1453 (73%) in both placebo and semaglutide group and 75% (1472) of the study population were white.
When you have an imbalance like this, it already makes it difficult to interpret findings and apply them to non white or female populations.
Another criticism I have with this trial, ( something to always look at when interpreting a paper is where did the funding for it come from?) This trial was funded by and designed by the drug company who make and sell the drug- Novo Nordisk. That in itself creates conflict of interest and potential biases. The randomisation was 2:1 this, although not uncommon in trials already results in unequal allocation and the potential for biases. Caution should be aired when reviewing confirmatory trials such as this. The trial was powered to be able to demonstrate a difference with these numbers, but if the hypothesis is that the group receiving semaglutide are going to benefit more than placebo, then really the justification for recruiting in an uneven fashion is null and void.
The next consideration that is not documented in the trial, what happens when people discontinue the medication or life style modifications? This study promoted a restrictive diet- calorie restriction. We know from a plethora of research that >95% of people who participate in diets regain the weight they lost and more, within 3 years. We do not know the long term changes in these participants. Additionally, the majority of the population within this trial were female and white and so I would be cautious to apply this to a wider population including non female or any other ethnicity.
I believe it is important to understand a few things before thinking about prescribing this weight loss intervention or seeking it. Fat does not equate unhealthy. Weight stigma and bias is unfortunately an inherent problem within healthcare and society and so there would be an obvious appeal to approve another drug to “fight obesity” . Being fat is not in itself a medical condition. The focus should be on addressing risk factors that are “associated with fat but are not proven causational” , and addressing those, for example cholesterol, high blood pressure and diabetes andit is likely more all of these are more nuanced than prescribing weight loss.
Prescribing weight loss seems like the easy solution. You’ll note, I personally do not adopt the terms, obese or overweight because these are both stigmatising and medicalising a size rather than an individual patient.
I am not proposing to have a solution, but adopting a one size fits all approach does not seem to cut it.
Further more, my intention was to highlight the evidence is not as black and white as you may hear and I’m imploring fellow health care professionals to entertain the notion that weight is not a reliable measurement of health in isolation. I encourage anyone who is not familiar with the Health at Every Size movement to educate themselves;https://asdah.org/health-at-every-size-haes-approach/. To not blanketly prescribe weight loss to patients in front of you. I am concerned, these medications have the capability to be abused, patients with eating disorders living in larger bodies are potentially at risk of being prescribed these rather than receive treatment for their eating disorders. As a doctor it is our duty to do no harm, please consider this when you have a fat patient sat in front of you.