top of page

Chapter 31: Math

Updated: May 30, 2023

cover of movie script

I’m a writer. I’ve always been a writer. I wrote a play in elementary school about an alien invasion that went undetected because it occurred during Halloween, a movie script in junior high featuring a flushing toilet as the whirlpool monster Charybdis in a remake of Homer’s Odyssey, and free-form poetry channeling Tyrone Green’s opus “Kill My Landlord” (played by Eddie Murphy on Saturday Night Live). As a general rule, writers don’t like math. At all. But somehow I found myself in an advanced calculus class in high school and I was struggling with a homework assignment. As in, I didn’t do it.

I was facing a no-doubter of an F, which is hard to do when the answers were clearly printed in the back of the calc book. All I had to do was show my work to pass. So I bent the rules as all high school juniors occasionally do: I copied down the question, added an equals sign, then wrote down the answer verbatim. If I got caught, I’d plead that it was so easy that I did all the work in my head, ergo no work to show, I was just that smart, Teach.

I got caught.

Fortunately I wasn’t the only one. One by one, my math teacher called students to his desk for a demonstration of their suddenly prodigious skills in mental gymnastics to solve calculus problems sans pen and paper. I clearly had one option to avoid the dreaded parent-teacher conference: double down on the ruse. In a panic, I scrambled to memorize the answers to as many calc questions as I could, notching all of three before I was called by the judge to approach the bench. Gulp.

My homework looked guiltily pristine. Long division is hard to do in your head. Calculus? My dad, a mathematician who taught the highest levels of math, couldn’t do it without at least a few chicken scratches. Mr. Kokonis knew he had me busted and covered up the first answer with his chalky hands—he speedily worked his equations on the chalkboard and rarely bothered with an eraser. “So David, what’s the answer?”

I paused, deliberately pondered for a few beats, then solved the question to the shock of the seasoned math teacher. He quickly covered another answer. “And this one??” Another pregnant pause before I birthed yet another correct answer. Stunned, he asked me to solve a third question. Aces. Now our battle of wits had reached its crescendo and my fraud was about to be unmasked. Question number four was my doom as I had memorized only three.

“Ohhhkay….” said a rightly skeptical Mr. Kokonis. He didn’t know how I did it and frankly neither did I. I’m far worse at memorization than I am at math. Case in point, when giving the best man’s toast at my friend Scott’s marriage years later, I chose the shortest quote possible to seal to memory, a mere seven words uttered by Mahatma Gandhi—“Where there is love, there is life.” I managed to get it backward, extolling to puzzled wedding guests instead how since Scott and his new bride Liz were indeed alive, there was love.

Twenty years later I dropped by my old high school and tracked down Mr. Kokonis, who incredibly was still teaching in his 70s, to tell him that story. He didn’t recall me or my hijinks… but he remembered Laura fondly, who was far brighter and one of his best ever students. His hands were black instead of chalky as the school had moved on to dry erase boards. I tracked him down once more researching this chapter and it was his turn to shock me: he was still teaching. At 90! And still doling out life advice to his pupils past and present. “I hope as they grow older that they will always remain positive and enthusiastic about whatever they are doing and that they will not be discouraged when things get difficult,” he said in a recent interview.

Uh, yeah, things for me had been difficult. Really difficult. And to get to the bottom of at least one of my pressing issues—this global Lyme disease kerfuffle—it required, of all things, analyzing numbers… and doing math.

Number Crunching

Days after I discovered I had Lyme disease and not multiple sclerosis, struggling to comprehend just how such an error could possibly have occurred, I had noticed the striking anomalies in the MS prevalence maps: the overwhelming tilt of cases being diagnosed in the far Northeast and upper Midwest, the heart of tick country. But how could I tease out viable, actionable information from this to help make my case? Even though Laura was the family’s scientist and math whiz, this type of data deconstruction required someone who does this stuff for a living, like an actual actuary. And where was I going to find an actuary? It’s not like there’s an actuarial Hinge, Tinder or Bumble app to swipe through to find a perfect match. Heck, I had a better chance of getting that Nobel before I found an AA (actual actuary). That is until July 13, 2022, at 1:16 p.m. when my email fortuitously dinged.

Hey Dave, so I see you are not pulling any punches. I guess it makes sense given the vehemence of your opposition, the last thing you'd want to do is blink. You asked for volunteers to help you and so that is what I am doing. I happen to be a Fellow of the Society of Actuaries with sophisticated data science skills. If you need any statistics crunched or estimates/projections made using full or partial data sets I would love to help. Perhaps my credentials could lend some credibility to any statistics supporting your case. Praying for you. Man, you're gonna need it. — Randall

I immediately swiped right and we connected—and by god, we couldn’t have been better suited for each other. We were both diagnosed with MS during the same year, we were both now wheelchair users having had similar disability trajectories, we both underwent aggressive stem cell transplants (only to later fail them), we are nearly the exact same age, and—although this detail is immaterial to this effort—we both share the same devastating handsomeness that transcends space and time. Well, I presume.

Although I didn’t need any further convincing, he sealed the deal thusly. “Just as background, I did the statistics [on my nurse practitioner wife’s] doctorate thesis regarding death rates for Oregon home births. There wasn’t much data to go on, but I was able to apply actuarial principles to predict a death rate 4 times as high as hospital birth. This was confirmed a few years later when new data became available (because my wife lobbied for changes to the death certificate forms).”

I sent Randall the seminal 2019 study (featured in Chapter 28: Hidden in Plain Sight) that extrapolated MS prevalence in the United States by examining health insurance databases—databases that are rich with data. Databases that show an extreme bias of MS in the far Northeast and upper Midwest, known tick sanctuaries. Within days a spreadsheet arrived via email with a cautionary note.

“Debating on where to draw the ‘what if’ line here,” Randall writes. “I have started at the most extreme of the possibilities with all of the ‘excess’ MS cases by region being Lyme. Clearly, that would be seen as inflammatory so I’m sure doing something more conservative would be the best approach.”

Conservative? Moi? But I agreed with Randall that maybe we shouldn’t spike our research with a beyond-inflammatory dumpster fire of extra Lyme disease cases just yet and play the “extreme” numbers straight. Insurance records pegged MS prevalence at 272.7 cases per 100,000 in the West, 272.6 in the south, 353.1 in the Midwest, and 377.4 in the Northeast, for a countrywide average of 309.2 cases per 100,000. To account for at least a smidge of cases picked up from travel like my own, Randall set the presumed true MS prevalence nationwide in his first pass at 270 cases per 100K, with a negligible amount of additional cases in the West and South. This would place the overall MS misdiagnosis rate in the United States at a conservative (for me) 12.6%.

The Excel spreadsheet Randall had prepared was overflowing with information, as he broke down MS by sex, age, and region, and then broke out Lyme cases by state based on CDC estimates before directly comparing the two. We got on the phone to discuss the results. They looked a little, uh, funny.

When assuming a scant 1% of cases diagnosed as multiple sclerosis in the West and South regions are in reality missed cases of Lyme disease, the spreadsheet spits out results that don’t entirely jibe in the South and Northeast. The new calculations suggest that doctors practicing in the tick magnet Northeast are certainly no better at correctly diagnosing Lyme disease than doctors in the South, and probably slightly worse. “Weird,” said Randall. “I would have thought that since so many docs see Lyme in the Northeast that they would have the lowest rate of misdiagnosis.”

Huh, me too. We then briefly did what researchers do when data doesn’t line up to a cherished hypothesis—we came up with hairbrained theories to rationalize why this discrepancy was entirely plausible. Maybe there was just so much Lyme in the Northeast that docs just couldn’t keep up and sufferers fell through cracks in the system. Or maybe physicians in the South were more hypervigilant because those rare Lyme cases really stand out from their typical workload of gator bites and injuries from wrangling 17-foot Burmese pythons.

Or maybe it was time to squirt into the mix some highly flammable lighter fluid and set that dumpster afire with a number that matched all my research: an ungodly, countrywide MS misdiagnosis rate of 30 percent.

One Third

The messy problem with science is that it needs to be repeatable. Consistently. And if a finding deviates from projections—if some expected outcome skews left instead of right—there better be a darn good explanation why it doesn’t line up like goosestepping North Korean soldiers at a missile parade in downtown Pyongyang. Countless eureka medical “findings” have been torpedoed when fellow researchers couldn’t replicate the results. That’s why for the last year I’d unintentionally been following the advice of Dr. Marshall, who recommended in a recent talk to budding young researchers—and apparently one old goober without a wit of medical experience who stumbled on said talk by happenstance—that they try to break their cherished theories. Think dinner plates at a Greek wedding, Opa!

Canadian Mounties

I recalled the first stomach-churning moment my Wedgwood teetered and threatened to make this memoir a moot mosaic. Do you remember “Jessica” from Chapter 17: Ipso Facto, the poor young woman who got the attention of the Canadian Mounties after she went on a psychosis-fueled crime bender? She was convinced it was Lyme disease and the docs were convinced she was crazy: it was just her MS. In that chapter, I discussed that true psychosis in MS is, in fact, rare, only 2-3 percent of cases. But what if the psychosis attributed to multiple sclerosis is solely from patients suffering from Lyme disease? That would make sense, as spirochetes have a long history of making people go cuckoo for Cocoa Puffs—all I had to do was look in a mirror for Exhibit A. Yet there was an issue.

If the prevalence of full-on psychosis in MS is a scant 2-3% of patients, and about one-third of patients diagnosed with MS actually have Lyme, that would mean the window of prevalence of psychosis in Lyme patients would need to be impossibly narrow to explain that the tickborne disease was responsible for the severe mental distress in reputed MS cases: just 6-9% (3 times 2-3%). Any percentage of Lyme psychosis remotely higher than that and my hypothesis flatlines, the math just wouldn’t work. For instance, if I discover that 30% of Lyme patients suffer from psychosis, the MS rate would have to be a third of that: 10%. I almost didn’t want to do the research. I had presumed that wackadoodle and Lyme were such frequent bedfellows that the rate would have been several times this estimated 6-9% range, at least.

Dave in a wheelchair next to melting metal sculpture
Phew, our dishes are safe.

A recent September 2021 article in Psychiatric News from the American Psychiatric Association allayed my concerns. “Despite the link between Lyme disease and mental disorders, the absolute population risk is low: just under 7% of patients in [a large] Denmark study diagnosed with Lyme in the hospital went on to develop new onset mental disorders requiring hospital contact.” The authors of the study went on to say that they believe the ultimate true rate of psychosis due to the tickborne disease is likely a touch higher, as there are no doubt less severe cases that did not result in hospitalization. So Jess, despite the over-the-top protestations of your neurologists, I believe you were right. The math supports the derangement that you, and a few of your MS-diagnosed friends, have experienced as being caused instead by Lyme disease. And thankfully, Laura could now rest assured that no China was going to be harmed by this latest revelation.

Time and again I’d ring the 30% bell, so much so that it was turning into a parlor trick. But could it stand up to broader scrutiny and provide insight into the bigger MS mysteries? One such puzzler surfaced in August of 2022 after a study was released with a confounding conclusion: The age at onset of relapsing-remitting multiple sclerosis has increased over the last five decades.

“Why?!” tweeted a bewildered Dr. Boster, an acclaimed neurologist out of Ohio. Dr. Beaber, the astute MS specialist out of So Cal, seconded his colleague’s head scratching. “This is despite the fact that easier access to MRI machines and changes in diagnostic criteria allow easier/earlier diagnosis. Mystifying.”

tweet from Dr. Beaber

All rational logic suggests the age of diagnosis should be decreasing, not increasing. Could Lyme disease cases mistaken as MS be the reason? After the study excluded patients with early-onset and late-onset disease that could significantly skew the numbers, researchers settled on a mean onset age of 28.38 years in the 1970-79 decade, an age that bumped up to a relatively grizzled mean of 33.37 years for the latest decade studied, 2010-2019. Counterintuitively, every decade, the age of onset grew a full year despite far better testing, technology, and diagnostics. In contrast, the average age of onset for neuroborreliosis, revealingly, is significantly older, in the mid to upper 40s. With studies for the onset of the neurological form of Lyme bouncing from ages 45 to 46 to 47, a mean age of 46 seemed appropriate to put my 30% misdiagnosis rate to the test.

“Laura, I need your math help, pleeease.” I knew when I was over my head in the math department and asked her to calculate what it would look like if 28-year-olds were comingled with 46-year-olds at a 70/30 clip, nothing else. Like the mathematics magician that she is (not to mention an incredibly patient spouse who is giving her husband time to right his relationship ship), she emailed me the answer straightway so I wouldn’t lose it among all my research… abracadabra alakazam!

Calculation: 70% x 28 years for MS + 30% x 46 years for Lyme = 0.7 x 28 + 0.3 x 46 = 33.4 years for the overall average age of diagnosis. Xoxoxo, Me

Presto. When geezerly Lyme patients are tossed into the jumble of younger, whipper-snapping MS patients, the 33.37 average age of onset for multiple sclerosis, rounded to 33.4, becomes fully explainable. Because it’s inaccurate—the age of onset for MS isn’t increasing after all. Thank you, Mr. Kokonis, for properly educating my brilliant wife.

Meanwhile, Mr. Kokonis’s advice to remain positive and not get discouraged was prescient. The CEO of University of New Mexico Health wrote back the morning after she received my decidedly unorthodox December email, aka Chapter 30.

email from UNMH

Mr. Bexfield, thanks so much for your letter. I've asked our Chief Medical Officer to

please follow up on this, and we will be back in touch. Thanks, Kate

There’s a reason I was so confident in those gaudy MS misdiagnosis numbers in the letter I sent her—an outrageous claim that at least 250,000 MS patients in the US and over 750,000 worldwide don’t have the demyelinating disease. That’s because Randall later reran those numbers at an appropriate dumpster fire-esque 30%. The first time I opened that spreadsheet I gasped. I expect you will, too.

Holy hell.

Dave leaning against a leaning fence

Read my Halloween elementary play:

Download PDF • 4.09MB


bottom of page