As my endo read through the medical chart at a recent appointment, I sat there anxiously waiting for him to tell me my latest A1C. He scanned the notes and rattled off bits of information about prescriptions, before getting to the meat and potatoes (so to speak) of our visit.
If you were a fly on the wall at that moment, you would have seen me all jittery, leaning forward in the chair waiting for the words to emerge. After what seemed like an eternity, he spoke:
Your A1C is 7.7%
My heart sank. While not really much different than my previous result, it was a 10th of a percentage point higher than last time at 7.6. Sure, it was only a tiny change, but in my head a screaming voice of judgement shouted: “Your A1C went up!” I was beyond bummed, especially because I’ve been putting in a lot of effort over the past few months to do better.
Apparently, this A1C was telling me that I was in fact doing worse than before, even if just a little bit.
Then I began to doubt… was this an accurate A1C result?
Since I decided in mid-May to take a break from my insulin pump, my blood sugars have actually been spectacularly better. I’m now using Lantus twice a day for basal, combined with NovoLog for fast-acting and Afrezza inhaled insulin for ultra rapid-acting bolus insulin doses. The goal is of course to increase the amount of time my BGs are in range (70-180), and reduce the number of big spikes and dips in BG levels. I’ve started seeing more in-range time since mid-May, and I’ve been happy with my success.
But I also reflected on how I’d been slacking some in the 2-3 weeks leading up to this particular appointment. My glucose variability had increased as I experienced more frequent higher BGs. So it was a bit of mixed picture.
Here’s what my CGM data shows:
Based on all of that, I believe that my most recent hemoglobin A1C result was somewhat “artificially inflated” from the highs in just the last few weeks — not reflecting the improvement I’ve been seeing in my diabetes management over the past three months in full.
In my opinion, this A1C result was lying to me and those who make decisions about my healthcare based on this number.
Science proves that this is a possibility…
To dig into this topic, we spoke with Dr. Irl Hirsch at the University of Washington, a fellow T1 and well-known researcher and expert on glucose variability, who has long criticized relying on the A1C as the gold standard of diabetes management. He confirms that it’s definitely possibly to “manipulate” an A1C with short-term changes, in the fashion that I suspected for mine.
The A1C “is a test you can study for,” Hirsch says. “The latest science shows that yes, even though your A1C is an average of the past three months, 50% of your A1C is based on glucose in the last month.”
He cites several studies on this, going back a decade and further — one of the more well-known being from 2008, when Dr. David Nathan found that recent glucose variability can impact an A1C result in people with T1D. The data from that study showed that the higher A1C levels, and among those with the highest glucose variability, the result could be off by as much as an entire percentage point!
Dr. Hirsch says that three studies now confirm the A1C doesn’t give the full picture of someone’s diabetes management trends. He points out that many things can impact an A1C result, from medications to iron deficiencies that can cause false A1Cs. Hirsch even notes that racial disparities exist in A1Cs, as scientific data now shows that for some reason in African-Americans, glucose binds more to hemoglobin and that can lead to A1Cs that are on average .3% higher than in Caucasian PWDs.
Hirsch says that someone with an A1C of 8.0% can have an average glucose ranging anywhere from 120 to 210.
“You’re basically throwing a dart,” he says. “We use this number to guide us on our diabetes, telling us whether it’s safe to get pregnant, the effect on complications, whether someone is ‘compliant’ or not, and now to determine how doctors get reimbursed. But it doesn’t show the entire picture, and you really have to look more closely at each patient.”
Ah ha, see?! We knew it!
Revisiting Standard Deviation
My endo agreed it was certainly possible my A1C was inflated, and in keeping with Hirsch’s work, recommend I also look at my standard deviation that measures how much your BG levels bounce up and down. (Reminder: low SD is good, high SD is bad, because it indicates big swings.)
It’s a bit of a confusing measure, but my doctor told me my deviation of 58 multiplied by 3 should be less than or equal to my average of 160 mg/dL, and mine came in just above that. So he described me as “borderline” but also said not to worry as I’ve been doing much better lately. That made me happy. It’s important to note that you can also have a good A1C level with poor standard deviation in diabetes. So complex!
This all backs up what many of us have been preaching for many years: A1C is not the end-all, be-all measure for evaluating diabetes care. A low A1C that traditionally indicated “compliance” doesn’t actually mean our blood sugars are staying in range as much as they should — and it certainly doesn’t take into account the dangerous Lows we may be experiencing. Same goes for the higher end of the scale.
On top of that, we’re more than just a number, and there need to be other, more meaningful measures of “success” with diabetes.
FDA Looks “Beyond A1C”
Fortunately, I’m not the only one thinking about this. The FDA is soon holding a day-long public workshop on
The meeting will delve into what the FDA should consider — beyond A1C impact — when evaluating new diabetes devices and drugs. This upcoming workshop follows the historic November 2014 webcast discussion between the FDA and Diabetes Online Community — the one where so many people tuned in live that we ended up crashing the FDA’s servers!
We’re delighted to see this finally being officially addressed, as I personally can think back to my younger days when I’d say to my diabetes care team: “I’m not doing this for a better A1C, I’m doing this so that I am not having severe High or Low blood sugars!” Quality of life with diabetes is about keeping things steady, after all.
To me, I think it’s important to look beyond A1C at three important data-points that are more reflective of how I’m doing:
- Time In-Range: This is a key measure for me, because this is an indicator of how on track my diabetes management really is.
- Hypos: These are dangerous and can lead to scary situations where I lose the ability to think and treat myself, and possibly even conciousness. If these happen overnight, I might not ever wake up. So the fewer Lows, the better.
- Glucose Variability: My blood sugars should be as smooth and steady as possible, since spikes and dips can lead to higher blood sugars and lows.
Personally, I just hope the FDA hears loud and clear: We are not just a number.
So, D-Friends: What do you want the FDA to know going into this workshop? What do you think needs to happen as far as broadening the view of positive outcomes with diabetes?