My training program consists essentiallyof the material covered in this book. It's my hope that physicianswho have little time to educate patients will use this book to assist in that purpose.
Establishinga Treatment Plan 115
SETTING A BLOOD SUGAR TARGET
Whenever I talkabout blood sugars in this book,I'm referringto finger- stick, plasma blood glucose measurements.
When I discuss "normal" blood sugar values, I am referring to those found in nonobese non diabetics — and to those not taken within 3 hours of a high- carbohydrate meal.
In my experience, given the right blood sugar meter, these values will be almost exactly the same as you would get from plasma mea surements of venous blood that your doctor would send to a clinical laboratory.
I've seen finger-stick blood sugars measured on hundreds of nondiabetic, nonobese adults (for example, salespeople who come into the office trying to sellme meters — I insist on demonstrations;* or the nondiabetic spouses, parents, or siblings of patients).It usually isabout 83 mg/dl.In order to simplify, I round offand tellmy patients that a normal to shoot for is 85 mg/dl, no matter what age.
I haven't had the opportunity to test a great number of nondiabetic children, but the literatureshows that normalblood sugars willbe about 85 mg/ dl, with the potential to be considerably lower.* With respect to hemoglobin A,c, I have a sophisticated machine in my office that I've found correlates almostexactly with measuresfrom a clinical laboratory.
I therefore check HgbAlc values on every patient at every routine visit, and frequently on nondiabetic relatives.
Essen tially what I see is that nondiabetics who are notobese have HgbAlc levels in the range of 4.2-4.6 percent.
I have a number of diabetic pa tients who, under treatment, now have HgbAlc readings aslow as4.2
*I usedto havesomefun withnondiabetic sales repswhentheycameinto the of fice selling bloodsugarmeters.
They'd be demonstrating a meter, which I would compare to my own meter.
I always used their blood because I've had enough finger sticks.
I'd "guess" their blood sugar.
I'd make a show of examining their skin, then give them a number.
It was always about the same, but they didn't know that.
The number was 83 mg/dl.
Inevitably I'd be within ±3 mg/dl.You know, of course, that I didn't have anyspecial powers — it wasjust that I'd seen so many random finger-stick readings from nondiabetics, I knew what number the nondiabetic waslikelyto show. t A study published in the New England Journal ofMedicine found that nondia beticmen with fasting blood sugars of 87 mg/dl or more had progressively in creased risk of developing diabetes than those with values less than 81 mg/dl.
Another study of about 2,000 healthy men, published in Diabetes Care in Janu ary 1999, showedthat overa period of twenty-two yearsthe risk of cardiacdeath was40 percent greaterfor thosewith fasting blood sugarsgreaterthan 85 mg/dl.
116 BeforeYou Start
percent. This is a considerable deviation from the ADA's recommen dation of under 6 percent — with no intervention unless levelsexceed 7 percent. In my opinion, this is yet another example of "the rape of the diabetic." The ADA recommendation for "tight control" of blood sugars, from its Web site, is as follows:
Ideally,this means levels between 90 and 130mg/dl before meals and less than 180two hours after starting a meal, with a glycated hemoglobin level less than 7 percent. The recommendations go on to state that tight control (what I advo cate) "isn't for everyone," which I believe is nonsense. But the ADA's tight control as definedabove isn't verytight at all.I would callit "out of control."
CONVERTING HgbA,c TO BLOOD SUGAR VALUES
Many years ago, I reviewed dozens ofHgbAlc values andthousands of blood sugars from data sheets submitted by my patients and came up witha formula forconverting HgbAlc to mean (average) bloodsugar.
Myformula doesnot jibewith most other formulas, perhapsbecause others haven't collected blood sugars throughout the day running into the hundreds or even thousands ofpatients covering4-month periods.
Theformula isvery simple.
An HgbA,c of 5percent isequivalent to an average blood sugar reading of 100 mg/dl,and every1 percent above5 corresponds to an additional 40 mg/dl increase in blood sugars.So an HgbAlc of 7 percent would correspond to an average blood sugar of 180 mg/dl.
The formula is, in myexperience, useless for HgbAlc values of less than 5 percent, and it may not work for average blood sugars greater than 300 mg/dl for the simple reason that for a new patient running blood sugars greater than 300 mg/dl, we rapidly get them down into the 100s or less.
Such new patients don't come in bringing me hun dreds of data points in the 300s for me to compute an accurate for mula at these values— nor would I ask them to.
Many may not bring me any prior blood sugar data on their initial visit.
In February 2002a study published in Diabetes Care reported a for mula that is valid for average blood sugars over a much wider range
Establishinga Treatment Plan 117
than mine, includingvalues well above and below 100 mg/dl.
It gives results close to mine in the 100-200 mg/dl range.
The formula is: mean plasma glucose = (35.6 x HgbAlc) - 77.3.
So how do we go about setting a target normal value given allthese numbers?
Let's take a look at a type 2 diabetic whose disease can be controlled by diet and exercise.
Here, we'll certainly shoot for blood sugars of about 83 mg/dl before, during, and after meals.
It will then be up to both me and the patientjointly— if his blood sugars are, say, in the 90s — to decide whether we want to introduce medications to further lower blood sugar.
Many patients these days are hesitant to take any medication that's been approved by the FDA, despite many suchmedications' beingquitebenign.
If we have atype 2 diabetic who requires the insulin-sensitizing drugs like metformin or the thiazoli- dinediones, we certainly canshoot for a target blood sugar of 83 mg/dl before, during, and after meals, and indeed, I will work with the pa tient to juggle the medications,usinglong- or short-actingversionsin orderto achieve that target.
Type 2 diabetics who require very small amounts of insulin (say, 1-2 units perdose) are at very low risk for hypoglycemia and will usu allyautomatically"turn off" the insulinthey make themselvesifblood sugars are too low.Such peopleare also goodcandidates for a target of 83 mg/dl.
When it comes to type 1 diabetics, wherevirtually allof the needed insulin is going to be injected, I increase the target to 90 mg/dl, even though we know that the mortalityrate — even in the general, nondi abetic population— is slightly greater for those with fasting or post prandial blood sugars of 90 mg/dT than it is for those with blood sugars of 83.If at all feasible without frequent hypoglycemic episodes, I will eventuallylower the target to 83 mg/dl.
I now use 83 as a target for myself.
A targetmay imply corrections to get you to your target.
As a rule, if you're a type 2, your blood sugar goes down eventually— maybe quickly,maybe over many hours.
If you'rea type 1 and injecting sig nificantdoses of insulin, if you make a mistake on your diet and your blood sugargoes up, you have to inject additional,calibrated doses of fast-acting insulin deliberately to bring down your blood sugar and, if it's too low, take glucose tablets to raise it.
For a new patient in the very early stages of type 2 diabetes, I may seeboth hypo- and hyperglycemia.
This is probablybecauseone ofthe early"lesions" of type 2 is difficulty in storing the insulin granules
118 BeforeYou Start
your body makes.
So such a person would make insulin for a meal, then make more after the meal.
A nondiabetic would store that addi tional insulin as it's being made, but the early type 2 would release some or all of it into the bloodstream as it's generated, thereby bring ing blood sugar too low.
This explanationalsoaccounts for attenuated (diminished) phase I insulin response—just not having enough in sulin stored to cover a meal adequately (another reason to follow a low-carbohydrate diet).
Such an individual could experience blood sugars in the 70s or even mid-60s from time to time, and these indi vidualsmust carry glucose tabletswith them to bring blood sugars up to their target, usually 83.
They don't take injected insulin to bring blood sugar down if it goes too high when they make a mistake, be cause their bodies willdo that for them, probably faster than injected insulin would.
SETTING GOALS OF TREATMENT
On the third visit, it's generally appropriate to prepare a fist of treat ment goals.
Exactly what are we going to accomplish, how, and over what time frame?
The patient and I discuss a list of goals to make sure that he or she understands and agrees.
The following list is typical of the things I want to see any given patient accomplish. (Remember, the training I provide to my patients is the substance of this book, so if you don't entirelyunderstand allof thesegoals right now,don't be dis couraged.Mark this chapter and come back to it when you'vefinished the book.
By then you should understand the whole philosophy of my approach and the goals will make sense.
You may also by that time have developed — if you haven't already— conscious goals of your own.)
• Normalization of blood glucose profiles. • Improvement or normalization of the following laboratory tests that respond to blood glucose control (Chapter 2): hemoglobin A1C thrombotic riskprofile red blood cell magnesium renal profile lipid profile • Attainment of ideal weight (where appropriate). • Full or partial reversal of diabetic complications, including pain or numbness in feet, diabetes-related retinal or kidney problems,
Establishing a Treatment Plan 119
gastroparesis, cardiac autonomic neuropathy, neuropathic erec tiledysfunction, postural hypotension, andsoon.
If bloodsugars are kept normal, some of these improvements willappear within weeks to years, depending upon the particular problem and its severity. • Reduction in frequency and severity of hypoglycemic episodes (where appropriate). • Relief of chronic fatigue and short-term memory impairment associated with high blood sugars. • Improvement or normalization of hypertension. • Reductionof demand upon betacells.
If C-peptideis present be fore starting our program (that is, if the pancreas is producing measurable amounts of insulin), glucose tolerance should im prove if a regimen is pursued that minimizes the demand upon the beta cells.
This is a very important goal.
Remember that for type 2 patients,smallsacrifices now can preventthe need for 5 or more daily insulin doses down the road.
Beta cell burnout (see page99) can frequently be prevented. • Increased strength,endurance, and feeling of well-being.
The patient may wish to add some personal goals.
The doctor should respect these if at all possible.
For example, I have several pa tients who arewillingto do whateverI ask,provided I do not put them on insulin.
I consider this a reasonable preliminary goal for some, even though it may increase the risk of beta cell burnout.
After all, if we cannot enlist a patient's cooperation,we achievenothing.
PART TWO
Treatment
The Basic Food Groups OR MUCH OF WHAT YOU'VE BEEN TAUGHT ABOUT DIET IS PROBABLY WRONG