In May of 2017, a few months after my second son turned one year old, I got shingles. When I visited a local clinic they ran a blood panel as part of the diagnosis and notified me that my a1c (a blood test commonly used to diagnose prediabetes and diabetes) was high. The normal range is below 5.7, and my levels were 6.0. Not off the charts, but enough to elicit concern.  I happened to have a physical with my doctor already on the books, so I asked them to replicate the test.

Three hours after I left their office the doctor called me up and asked me if I was sitting down. Then he explained that my blood sugar records had been rising, from 5.4 in 2014 to 6.0 in 2017, and that this meant I had prediabetes heading into diabetes. “I’ve seen this a million times,” he said. “You’re in great shape, eat well, and take care of your health. You’re just genetically predisposed.”

The doctor explained that we should schedule a follow up appointment in a few months to get me on Metformin, a medicine used to treat type 2 diabetes, and that I should expect that in a couple of years I’d need to move to injected insulin and finger prick tests several times a day. Potential complications from the disease could include reduced lifespan, amputation, and blindness. 

When I asked what I could do to revert the condition the doctor paused before saying, “I’ve seen hundreds of cases like yours. They all resulted in Type 2 diabetes.”

Diabetes is a collection of interrelated syndromes stemming from poorly managed blood sugar levels. Its causes, severity and impact are affected by our environment, our diet, our exercise, our stress and a huge number of other factors. Since my diagnosis I’ve met people who get terrible blood sugar spikes from bread but not candy, rice but not juice, or skipping a workout but not excess sleep. In other words, It’s hugely personal. I do just fine on oatmeal, for example, but white rice makes my blood sugar spike so badly I get headaches.

This makes diagnosing the “cause” just as hard. Of course genetics can play a role, but so do all of the factors mentioned above.  I’ve met people with diabetes who run marathons, and people with extreme obesity who have excellent blood sugar. 

It’s also enormously pervasive. Approximately 96 million Americans – or roughly 1 in 3 – have prediabetes. Those with prediabetes have a 50% chance of contracting full blown diabetes within 5 to 10 years.  Once diagnosed, the condition is generally considered irreversible.  What’s more chilling, 9 out of 10 people who have prediabetes don’t know it and therefore are not actively addressing it.

Because diabetes (and prediabetes) are commonly associated with excess body fat, Americans seem to shy away from talking about it. In my case I was lucky, obstinate, and scientific – but mostly lucky. After my doctor called and explicitly doomed me to becoming diabetic I went on a warpath of self-education and experimentation, reading everything I could about prediabetes, diabetes, and possible treatments. 

That isn’t a luxury everyone has the time for, and I was extraordinarily lucky that my wife, my work, and my education all supported my inquiry.  I’m also a data guy – I’ve worked and played in emerging technology since my early school days and have been fortunate to work with world leading tech experts in a variety of fields. Moreover, when all this happened I’d just founded a data analysis company, so I was uniquely positioned to ask these questions.

What I found was more or less the above; that prediabetes is complicated, that it’s personal, and that there are a huge number of factors which can improve or reduce your chances of remission. 

But remission (from prediabetes, at least) IS possible. At the time, it was estimated that 10% of people who are diagnosed prediabetic revert from the diagnosis. I resolved to be in that 10%, and began by figuring out what I could test and control. 

I started by ordering a Continuous Glucose Monitor (CGM). While these were hard to get at the time, they’re much more common now, and a number of startups now offer them as lifestyle accessories to help people achieve weight loss and fitness goals. The critical part for me was being able to objectively measure the results of my self-tests; In other words, I needed to know for certain how various changes were positively or negatively impacting my blood sugar levels.  

The second thing I did was research factors that had shown, via published scientific papers, to improve insulin resistance and lower blood sugar levels in some cases. Then I made a list of all the ones I could control. Some examples included:

– Supplements (Cinnamon, Turmeric, Fenugreek, etc.)

– Specific food reduction/removal (dairy, carbs, gluten, etc.)

– Lower carbohydrate intake

– Ketosis (no carbohydrate intake)

– Exercise

– Stress reduction

– Sleep

Then I drew up a protocol for each test and got to work. Along the way I learned a ton, both about my body’s response to different protocols but also about how to run self tests. For example, if you want to determine how to lower your blood sugar, don’t go into ketosis, start taking a ton of supplements, and begin meditating all at once. You may get results, but you’ll have no idea why. 

Similarly, time and again I found that my results were intensely personal. When I tested a series of specific foods (“eat one cup of white rice after waking, measure blood sugar at 10m, 30m, 60m, and 120 minutes”) I found that some foods which I expected to cause a huge spike didn’t, and others did. Some reliably caused spikes unless I ate them with something else – I tested toast three separate times because I put butter on it the first time and got almost no sugar spike – until I skipped the butter. 

But mostly I learned that getting clear results was difficult because there are so many overlapping impacts on blood sugar. Eat right, exercise well, avoid stress, and if you happen to sleep badly you could wake up with exacerbated high sugar levels (the “dawn effect”). Ultimately, I realized that not only did I need to test one thing at a time, but that I needed to test one thing at a time for a long time at a stretch in order to obtain a useful representative average. 

Even then, it took some extraordinary circumstances to help me turn things around.

One of the weirdest things about my experience with prediabetes was that there were so many mitigating factors – but ultimately that proved to be the key. As miraculous as the devices are, running my phone over my CGM would give me a single number every time: my blood sugar level. Given that I was wearing it because those levels were too high, scanning my phone became an exercise in terror. Worse yet, I started realizing that – to a degree – I could act (in the short term) on these results. 

That meant that whenever I scanned my CGM I would either a) breathe a sigh of relief that it wasn’t worse than the last time, or b) panic. Mostly I panicked. And then, curiously, the number would go up.  The anxiety that I had over worrying about my blood sugar was actually making my blood sugar rise.

Eventually I took to scanning the CGM and showing it to my wife without looking at it. If she shrugged and rolled her eyes, I’d ignore it. If she grimaced, I’d drop to the floor and do push ups. It was only at the end of the day that I’d download all the data and look for trends, which is good, because oftentimes your blood sugar is reacting to events that happened hours before, be it food, exercise, or sleep. 

That proved to be the key. Over time, as I cleaned up my testing protocols and started rejecting controls which weren’t working, I began to realize that there was one primary driver of my own personal blood sugar levels: my stress. The final blow came when I decided to enjoy some ice cream with my extended family at a barbeque, and in playing with the kids ended up scarfing down several bowls – with no measurable impact on my blood sugar whatsoever. But an hour later, when my landlord called with a (thankfully inaccurate) bed bug scare my levels screamed past normal and solidly into kidney-damaging territory – and stayed there for hours. 

After some additional experimentation and consultation with doctors we came to the conclusion that – again, in my specific case – my body responded to stress by dumping a lot of sugar into my bloodstream. This was a problem because I’d optimized my career for stress; public speaking, startups, executive consulting, etc. But knowing is half the battle, and once I realized that I could eat my ice cream as long as I wasn’t worried about anything I started making lifestyle changes that had a functional impact. 

I won’t bore you with what worked for me here, but suffice to say that after some meditation, therapy, and a significant reshuffling of clients (sorry, angry entitled people!) my blood sugar levels settled and never rose again. I’d cracked the code.

It can’t be overstated that I was extraordinarily lucky, and that my case isn’t like your case, or anyone else’s. Again, prediabetes and diabetes are complex collections of interrelated syndromes. My doctor wasn’t lying when he said he’d seen cases like mine before which reliably resulted in Type 2 diabetes no matter what the patient did. 

But I’m also living proof that with attention to detail, and some careful analysis, sometimes people can revert prediabetes.  It also underscores what mindful experts and some nutritionists have been saying for years: we’re just starting to understand the unique and complex interplay of diet, stress and exercise on our health. If this worked on my prediabetes, what other conditions might we be able to rethink if only we understood this better? In my case it also meant less fat around my middle, a significantly improved understanding of what I can comfortably eat (goodbye, white rice!), and a hell of a lot less stress – all of which are good things. For every body.