The measurements that show up on your chart
Mortality is the outcome that matters most, but it is not the one you see move month to month. What you actually see at appointments are numbers: HbA1c, fasting glucose, weight or BMI, blood pressure. This part is about what happened to those after six months on the combination.
Four of them moved in a way that was statistically significant. One did not.
HbA1c dropped by 1.48 points
HbA1c is your average blood sugar over roughly the last three months, expressed as a percentage. It is the number most diabetes decisions revolve around.
Across seven studies and 1,011 people, starting the SGLT2 inhibitor and GLP-1 receptor agonist together lowered HbA1c by an average of 1.48 percentage points at six months (95% CI 1.21 to 1.75).
To make that concrete, most of the people in these studies started with an HbA1c somewhere between about 7.9% and 9.1%. A drop of around 1.5 points takes someone from 8.8% to roughly 7.3%. For many people that is the difference between "your diabetes is poorly controlled" and "you are close to target".
That is a big move for any diabetes treatment.
Fasting glucose fell by 2.27
Fasting plasma glucose, the number you get from a finger prick first thing in the morning before eating, dropped by 2.27 across five studies and 945 people.
One thing worth flagging: the paper reports this without stating the unit in the result itself, and the included studies came from several countries. Readers used to mmol/L will read 2.27 as a very large drop. Readers used to mg/dL would read it as tiny. Given the size of the HbA1c change, mmol/L is the plausible reading, but it is exactly the kind of small ambiguity you should notice. Always check the units.
BMI down 1.71
Body mass index fell by 1.71 points across four studies. For someone of average height that works out to something like four to six kilograms. The average starting BMI in these groups ran from the high 20s to nearly 40, so this is a meaningful shift for people who are carrying significant excess weight.
Part 1 explained why this pairing is good for weight in particular: the SGLT2 drug burns calories through your urine, and the GLP-1 drug stops your appetite from racing to make them back.
Systolic blood pressure down 6.35
Across five studies, the top number on a blood pressure reading fell by 6.35 mmHg.
That might not sound dramatic. It is. Blood pressure reductions of around 5 mmHg are the kind of change that, across populations, meaningfully shifts rates of stroke and heart disease. A lot of people take an entire extra blood pressure pill to get that much.
The paper credits most of this to the SGLT2 side. These drugs cause natriuresis, the loss of sodium in urine, which takes pressure off the system. A previous review the authors cite found SGLT2 inhibitors were more potent at lowering both systolic and diastolic pressure than GLP-1 drugs or DPP-4 inhibitors. The GLP-1 side adds its own help through effects on the kidney and blood vessels.
The one that did not budge
Total cholesterol went down by an average of 0.11, but the confidence interval crossed zero (from 0.25 lower to 0.02 higher) and the p value was 0.09. That means it was not statistically significant.
I think this is actually a useful result. It tells you the combination is not a cholesterol treatment. If your lipids need work, that remains a separate conversation, usually about statins. Knowing what a treatment does not do is as valuable as knowing what it does.
Putting the four together
Stand back and look at the pattern. Sugar down. Weight down. Blood pressure down. All at once, from two medications, without the headline risk that used to come with pushing glucose lower.
That last point is important. Older diabetes drugs that forced blood sugar down aggressively, like insulin and sulfonylureas, carried a real risk of hypoglycaemia, dangerously low blood sugar. In the studies here, symptomatic hypoglycaemia was reported in under 10% of patients in one study and mild hypoglycaemia in under 5% in another. Neither drug class pushes insulin out regardless of glucose level, which is why the paper describes the pairing as working at low risk of hypoglycaemia.
The adverse events that did show up are the ones already familiar from each drug alone: genital yeast infections and urinary tract infections on the SGLT2 side, nausea, bloating and diarrhoea on the GLP-1 side. In the Korean study, stomach side effects were common at three months but faded at later check-ins. Ketoacidosis, pancreatitis, fractures and acute kidney failure were not reported.
In Part 5 we ask a question that sounds small and turns out to be practical: does it matter which drug you start first?
This series discusses a peer-reviewed meta-analysis of observational studies. Nothing here is medical advice. Do not start, stop or combine prescription medications without your doctor.
