Claim That GLP-1s Cause Blindness Debunked

The internet has found its latest reason to justify moralistic attitudes about GLP-1 treatments: claims they cause blindness. Lyle McDonald decided to examine what the actual data shows in a recent YouTube video.

The controversy stems from a 2024 JAMA Ophthalmology paper examining the risk of non-arteritic anterior ischemic optic neuropathy (NAION) in patients prescribed semaglutide. NAION is the second most common form of optic neuropathy and a significant cause of blindness among adults. The study was motivated by anecdotal clinical observations from ophthalmologists who noticed a potential link.

The research examined databases comparing type 2 diabetes patients and those with obesity who either used or didn’t use GLP-1s. Among 710 type 2 diabetes patients, NAION occurred in 17 patients using semaglutide versus six in the comparison group. Among overweight or obese patients, the condition occurred in 20 semaglutide users versus three non-users.

While these numbers might seem alarming when framed as “seven times more,” McDonald emphasizes the need to examine absolute terms. The European Medicines Agency concluded that NAION is a very rare side effect, affecting up to one in 10,000 patients. Large epidemiological studies estimate semaglutide use is associated with just one additional case per 10,000 person-years.

The American Optometric Association’s 2025 clinical report addressed concerns about diabetic retinopathy complications observed in the SUSTAIN 6 trial, where 3% of semaglutide users experienced complications versus 1.8% on placebo.

However, McDonald points out this was “likely due to the rapid HbA1c reduction, a known contributor to early worsening of retinopathy.” This represents a transitory effect of improving blood glucose control in diabetics, not a direct effect of GLP-1s on the eye.

The November 2024 FDA announcement requiring safety labeling changes sparked concern. However, McDonald urges readers to look beyond headlines.

The FDA’s meta-analysis of 126 randomized controlled trials involving 134,253 patients identified 15 probable NAION events: 10 in patients treated with GLP-1s and five in patients receiving placebo.

The incidence rate was 6.4 events per 100,000 person-years for GLP-1 users compared to 3.8 per 100,000 for comparator-treated patients. McDonald calculated this represents a 0.01% risk. For perspective, the incidence of NAION in the general population of adults over 50 is 2.3 to 10.2 cases per 100,000 persons per year.

“This is 10 people out of 78,000 users of GLP-1s,” McDonald states. “Yeah, it’s a doubling from three to six per 100,000 person-years. The other one, one more person at risk for this.”

He acknowledges blindness is serious and any risk factor needs consideration, but argues the response online has been disproportionate. “You are just looking for an excuse to justify your moralistic negative attitudes against a treatment that is having massive benefits.”

Even if the risk were a full 1%, which it’s not, it would need to be weighed against the documented benefits for diabetics achieving better blood glucose control and people with obesity losing weight. McDonald concludes: “0.01%, 10 versus five people across 126 trials and 75,000 people. Guys, just stop.”