Trials on GLP-1 drugs and blood pressure measure changes after treatment starts. This national survey compared people at one point in time. It found differences that a randomized trial would need to put in context.
What the survey counted
The analysis used NHANES data from 2017 to March 2020. It included 4,016 U.S. adults with hypertension. Of those, 45 reported using a GLP-1 drug [1].
The estimated share below both 130 systolic and 80 diastolic was 66% among users and 24% among nonusers. After applying survey weights, the 45 users represented an estimated 994,028 adults[1]. The weights account for how people were sampled. The precision still depends on the number of respondents and the survey design [2].
Why the groups are hard to compare
The weighted share with diabetes was 99% among users and 18% among nonusers. Average BMI was also higher among users, at 37.25 versus 31.22 [1]. The groups differed substantially before considering blood pressure. The survey cannot show whether the difference came from the drug or other factors.
What randomized trials measured
A separate analysis used individual data from three semaglutide trials. Systolic blood pressure fell about 5 mmHg more with semaglutide than with placebo [3]. That measures the average change in systolic pressure. It cannot tell us how many people reached the survey cutoff.