I posted this to Reddit, but I figure it deserves to be here too.
We don’t actually know how many people have lupus, and I am confident this is harming patients:
Since being diagnosed, lupus has become one of my special interests (I’m autistic, so when I say that, I mean it in the full-blown “I have read every paper I can find and have built many SARD specific tools” sense). And something keeps happening: the more I learn, the more lupus I find. In my own family: three generations of women on my father’s side, all in central Oklahoma, all with SLE. Among coworker’s family members or themselves. Among friends and friends of friends. Some of my clinicians. Everywhere I look, lupus, or some flavor of it (MCTD, DM, Incomplete lupus, cutaneous, UCTD.. you get the drift) seems to exist in one way or another.
For a while I wondered if this was just the Baader-Meinhof effect, you buy a red car and suddenly every car on the road is red. But then I started reading how SLE prevalence is actually measured in this country, and I realized: it might not be that I’m suddenly noticing lupus everywhere. It might be that lupus actually is everywhere, and we’ve just never properly counted.
I’m posting this because of the recent thread here about racial prevalence of lupus after that interview between those two people I’ve never heard of were posted saying incorrect things about lupus prevalence. There were some good points made in the thread, but I think the conversation was missing a critical piece: the numbers we’re all arguing about are far less solid than most people realize, and that uncertainty isn’t just academic; it’s shaping how our doctors and society views us, and ultimately shapes health outcomes down the line.
Here’s the situation:
SLE is not a nationally reportable disease. Unlike cancer, TB, or HIV, no one is required to report new lupus cases to any central database. There is no national lupus count. What we have instead are estimates, built from a small number of regional registries.
The number most commonly cited in research, about 204,000 Americans with SLE, comes from a 2021 meta-analysis (Izmirly et al., Arthritis & Rheumatology) that pooled data from five CDC-funded registries. Those registries covered portions of Georgia, Michigan, California, New York, and the Indian Health Service. The data was collected between 2002 and 2009, then extrapolated to the 2018 census. The authors themselves describe this as likely representing a lower bound.
The older “1.5 million” number the Lupus Foundation uses? That comes from a 1994 telephone marketing survey.
Neither of these is a count. They’re estimates built on estimates.
And when a different methodology is used, a nationally representative population survey (MEPS, 2016–2018), the number roughly doubles. That study found a prevalence of 195 per 100,000, which translates to around 490,000 adults with SLE.
Why does this gap exist? I propose several reasons:
The registries used ACR classification criteria to define cases. These criteria were designed for research classification, not clinical diagnosis. The Lupus Foundation itself has noted that at one major academic medical center, only half of patients actively receiving lupus treatment met those criteria at that time, or ever. So the registries were structurally designed to miss roughly half the population being treated.
The registries also excluded cutaneous lupus, incomplete lupus, and drug-induced lupus entirely. And they covered parts of five states, not the country.
Why this matters for you, personally:
When a doctor “knows” that lupus is rare and primarily affects young Black women, that shapes how they evaluate every patient who walks in. A white woman with fatigue and joint pain gets worked up differently than a Black woman with the same symptoms. An older patient gets told it’s “probably not lupus.” A man doesn’t even get considered.