I was halfway through an intraoral massage when my facialist asked me when I last had my iron levels checked. My under-eye bags were looking a little blue (which, in retrospect, was likely ancient filler that had migrated) and my hair was seemingly malting (the perils of being a werking mother). So I went ahead and booked an all-encompassing blood test with my doctor.
Five days later, I got the results and alarm bells started ringing. At 203.73 µg/l (micrograms per liter) my ferritin—the protein that stores iron—was not low, but rather, raised. Anything over 150 is considered above the normal range. My vitamin B12 was elevated too, coming in at 1,669 ng/l (nanograms per liter), over double the normal level.
Help! Was I dying?
As a chronic procrastinator—welcome to the realities of having ADHD—I didn't get around to reading my results until the weekend, when, of course, my doctor was offline. And so I did what any neurotic millennial would do: consulted Dr. ChatGPT.
According to a study done by OpenAI, 40 million ChatGPT users ask the chatbot about health every day. In fact, more than five percent of all prompts fed to ChatGPT are about health, be it about symptoms, medications, or treatment options. This suggests two things: we’re all raging hypochondriacs, and ChatGPT is fast becoming our go-to health ally. With its ability to demystify complicated medical jargon, not to mention its 24-hour availability (very important for someone as codependent as me) and soothing tone of voice, I can see why.
“Why are my ferritin and vitamin B12 levels so high?” I wrote frantically, giving absolutely zero context about the rest of my health. “Is there something terribly wrong with me?” Then came the reply: “A combination of elevated ferritin and elevated B12 can sometimes point toward: liver disease, chronic inflammation, certain blood disorders affecting blood cell production, and, less commonly, some cancers. It can also be associated with heavy alcohol use.” Oh god, I thought, looking down at the glass of wine in my hand, so I am dying. More panic and questions ensued: “Do I have liver disease? Am I ok?”
I was on the precipice of a mental breakdown while researching various—ahem—wellness facilities to check myself into when ChatGPT came back with a few questions of its own, asking for my age, sex, and whether I was taking any supplements, multivitamins, or injections. That’s when it dawned on me that I was, in fact, taking an industrial amount of supplements and, just two weeks prior, had an intravenous vitamin infusion, the contents of which I was unsure of. No one wants to know what goes in the sausage. But come to think of it, there was some mention of vitamin B, which is said to be great for energy levels as well as regulating the nervous system, both of which I was in dire need of fixing.
As for the ferritin level, my chatbot reassured me that 203.73 µg/l was only slightly elevated, and probably just pointed to a mild case of inflammation. But it didn't stop there. Keen to get to the bottom of what was bothering me, Dr. ChatGPT asked a new series of questions. How was I feeling otherwise? What was my diet like? How many units of alcohol did I consume each week?
That's when the floodgates opened, and I started entering intimate details about my health, from my lazy gut and spartan poops, to the odd cigarette on the occasional boozy weekend. What followed was a 10-minute back-and-forth between me and Dr. ChatGPT.
At the end of our session, we settled on the following: my vitamin B12 levels were probably raised due to my recent IV drip, while my gut was potentially irritated by my aggressive supplement intake, food sensitivities (all those late-night pizzas), smoking, alcohol, and lack of sleep. To counter all this, ChatGPT suggested I reduce my alcohol intake, stop smoking, simplify my supplements, stay hydrated, and prioritize fiber-heavy meals—but not before advising I check with a real-life doctor.
“ChatGPT currently sits outside medical device regulation,” says Dr. Catherine Shuster Bruce, a human doctor and editor at medical publisher BMJ Group. That “means we shouldn't be using it to diagnose a medical emergency or guide treatment decisions. It could delay life-saving care.”
That said, a lot of people find it helpful. Most people use it when they first feel sick, ahead of a clinic visit, or to unpack medical lingo. “Its value depends on many factors including: how you're using it, what you ask it, where you are, and who you are,” says Shuster Bruce. Regardless, she advises caution. “ChatGPT is a language model that predicts the next most likely token, a fragment of a word, based on probability. It can validate our ideas, sound confident and also make stuff up — 'hallucinate' — generating plausible-sounding, false information.”
In my case, ChatGPT helped me off the ledge, saved me from a sense of impending doom, by contextualizing my data and asking insightful prompts, which proved extremely helpful when I did eventually speak with my doctor. Which is the whole point. ChatGPT can be an excellent aid when dealing with opaque medical information and to help shed light on early-onset symptoms, but it should always be cross-referenced with a clinician.
In the meantime, it is just quite nice to have someone to talk to.
