Diabetes IV: Continuous Glucose Monitors (CGMs)

/

Background 

  • What is a continuous glucose monitor? 

    • CGM - a wearable device typically put on the back of the arm, stomach, or buttocks that is composed of a sensor and a transmitter 

      • The sensor is a small needle/probe that sits in the skin and measures interstitial blood sugar 

      • It typically will clip into a transmitter that can then send this information to a device (ie. via bluetooth to your phone or to a website that your physician can check) 

      • Sometimes, it requires scanning the transmitter with a phone or another device to show the blood sugar 

    • Some logistics 

      • Typically can be worn for 7-14 days 

      • Should be active >70% of the time 

    • Other cool things it can do 

      • Also, certain CGMs can sync with insulin pumps to help better regulate blood sugar (ie. closed-loop insulin technology) 

  • Who will you see that has a CGM? 

    • Most likely patients with T1DM - insurance is more likely to cover 

    • Some patients with T2DM, though much less common 

    • Now, it seems that more patients with GDM who are not able to do fingersticks may obtain or desire a CGM.

    • Glycemic targets in pregnancy

      • Remember that the ADA and ACOG recommends targets for fasting of <95 mg/dL, 1 hr postprandial <140 mg/dL, and 2-hour postprandial of <120 mg/dL  

What are the numbers I should be looking at in a CGM? 

  • The targets can be confusing because instead of just 4 time points, we now have many, many more! 

    • Many CGMs will sample blood sugar every 5 minutes 

  • Things to look at should be glucose targets 

    • Targets can be individualized, but in pregnancy, the target should be between 63-140 mg/dL per the ADA 

    • Can ask patient to generate a report for you or when you log into their reports, you can generate a report for the last XX amount of days 

    • Some people will spend some time both above and below target (note that some patients’ targets may be individualized and different) 


Is a CGM actually useful in treating diabetes? (ie. does it improve outcomes?) 

    • Multiple studies done in patients with T1DM and pregnancy 

  • Largest: Continuous Glucose Monitoring in Women with Type 1 Diabetes in Pregnancy (CONCEPTT) - multicenter RCT that compared self-monitored blood glucose + CGM to SMBG alone in 325 women who were either planning pregnancy or who were pregnancy 

    • Those with CGM had a small but statistically significant difference in A1C (-0.19%) 

    • Those with CGM in pregnancy had statistically significant less time above range than control, without an increase in time below range or in number of severe hypoglycemic episodes 

    • Also there were differences in neonatal outcomes! 

      • In CGM group, there was lower incidence of:

        • LGA

        • Neonatal hypoglycemia,

        • NICU admissions

  • What about T2DM? 

    • Outcomes here are less robust, and there are fewer studies - no RCTs have specifically studied T2DM in pregnancy with CGM to date 

  • GDM?

    • Largest study was prospective cohort in 2014 - 340 women with GDM who had 4 weeks of blinded CGM + standard care or standard care alone

      • Those blinded to CGM had lower rate of preeclampsia, improvement in CGM metrics, and lower rate of neonatal composite outcome (ie. premature delivery, macrosomia, LGA status, SGA, obstetric trauma, neonatal hypoglycemia, hyperbilirubinemia, and respiratory distress)  

    • Studies also show that those with CGM are more likely to receive insulin therapy 

  • So some conclusions: 

    • Clearly, CGM can improve outcomes in T1DM and those that are pregnant

    • Less clear in those who have T2DM and GDM 

How do I manage those with CGMs? 

  • This should be done with endocrinology or MFM 

    • The goal is to get the patient at least 70% into the target range as described above

    • We won’t go into pumps  

  • For those with injectable insulin

    • It is still helpful to look at when the patient is having spikes in blood sugar and adjust based off of that 

  • If there are spikes after breakfast or dinner → add fast acting or regular insulin at those times 

  • If there are spikes with lunch → can either add fast acting with lunch or increase long acting in the morning 

  • If there are spikes with fasting, look at overnight glucose 

    • If there is a drop in glucose in the middle of the night (ie. 4 am) and then an increase, this is known as the Somogyi effect 

    • The way to address this is either (1) to decrease night time long acting insulin or (2) to add a protein snack after dinner 

  • If overall blood sugar is elevated in the AM between meals, can increase AM long acting insulin 

Updates on Gestational Diabetes Screening

/

We last talked on the podcast about gestational diabetes with Dr. Coustan very shortly after we began in December 2018. Those podcasts are so good, and feature the man himself who is co-credited with the “Carpenter-Coustan” criteria we all know…

But if you’ve been watching journals recently, you probably have seen a lot of interesting papers with respect to GDM screening. So today’s episode will be a bit of an update on part one of those past GDM podcasts! Treatment fortunately hasn’t changed much so we won’t update that part today.

By the way – ACOG PB 190 on Gestational Diabetes Mellitus is still an excellent read!

Physiology of Insulin Resistance in Pregnancy 

  • Progesterone effects on insulin resistance

    • Normally, insulin binds to insulin receptor → phosphorylation of beta-subunit of receptor and leads to phosphorylation of the insulin receptor substrate I (IRS-I) 

    • Progesterone reduces expression of IRS-1  

  • Human placental lactogen effects on insulin resistance

    • Both insulin-like and anti-insulin effects 

    • Generally decreases maternal insulin sensitivity

    • Decreases maternal glucose utilization

    • Increases lipolysis and free fatty acids

      • This allows for free fatty acids to become available for mother’s metabolism (do not cross placenta) so fetus gets glucose preferentially

  • On the fetal side, exposure to hyperglycemia:

    • Leads to increase in its own endogenous insulin production and production of insulin-like growth factor 1

      • These lead to increased growth, fat deposition, and risk for macrosomia.

Prevalence and classifications of diabetes 

  • Prevalence - hard to know b/c not everyone is tested

    • 2009: 7% of pregnancies were in people with diabetes.

      • 86% of these cases were GDM  

  • Classification: The White Classifications

    • Depending on where you are, you may still see these in use; they are also helpful in classifying pre-gestational diabetes.

    • Named for Dr. Priscilla White, who developed the schema in the 1950s and 1960s

      • Class A1: diet-controlled GDM

      • Class A2: medication-controlled GDM

      • Class B: onset at age 20 or older or with duration of less than 10 years

      • Class C: onset at age 10-19 or duration of 10–19 years

      • Class D: onset before age 10 or duration greater than 20 years

      • Class E: overt diabetes mellitus with calcified pelvic vessels

      • Class F: diabetic nephropathy 

      • Class H: ischemic heart disease

      • Class R: proliferative retinopathy

      • Class RF: retinopathy and nephropathy

      • Class T: prior kidney transplant

Complications of GDM - ie. Why do we care? 

  • Maternal complications

    • High risk of developing preeclampsia, undergoing C/S

    • Increased risk of developing type 2 diabetes later in life (up to 70% of patients with GDM develop T2DM within 22-28 years after pregnancy) 

  • Fetal complications

    • Increased risk of macrosomia, neonatal hypoglycemia, hyperbilirubinemia, shoulder dystocia and birth trauma 

    • Increased risk of stillbirth 

    • Fetal exposure to maternal diabetes may contribute to adult-onset obesity and diabetes in offspring 

Screening for GDM - The Basics 

  • Used to be medical history and past obstetric outcomes and family history → fails to get 50% of patients with GDM 

  • All pregnancies should be screened between 24-28 weeks of gestation, with one of two strategies:

    • Two Step:

      • 1973: O’Sullivan et al - described a 1 hr GTT with 50g of glucose  

        • Carpenter-Coustan criteria: positive if >/= 130 mg/dL, though some institutions use 140 mg/dL

      • If screened positive, should get a follow up test with 3hr GTT (100g)

        • Diagnose if 2 abnormal values 

          • However, even 1 elevated → increased risk of adverse perinatal outcomes compared to those without GDM or elevated values

      • Carpenter-Coustan: Fasting: 95, 1 hr: 180, 2 hr: 155, 3 hr: 140  

      • Alternative criteria (not in wide use): National Diabetes Data Group: 105, 190, 165, 145 

    • One-Step alternative screening: International Association for the Study of Diabetes in Pregnancy Group (IASDPG) method:

      • 75g 2hr test - just one test!

      • Fasting: 92 mg/dL, 1hr: 180 mg/dL , 2hr: 153 mg/dL 

      • If > 1 elevated = GDM

  • “Early GDM screening” to look for early gestational diabetes  should be considered in some patients with risk factors.

    • The best test to use for early screening is up for debate, however:

      • Some might consider A1c, but because of new red cell generation / faster turnover in pregnancy, may artificially lower the A1c

      • Some consider using an OGTT, but then it might be hard to convince patients to do it again if they screen negative.

      • Some might consider a trial of “glucose profiling” with a glucometer but not any rigorous testing done about this.

ACOG PB 190


Updates in the World of GDM Screening:

  • What’s better: two-step Carpenter-Coustan style, or one-step IASDPG style?

    • In the last year, two randomized trials (NEJM, AJOG) and a systematic review/meta analysis (Green) have been published to help answer this question. 

    • Because findings are similar, summarized from the meta-analysis:

      • Patients with one-step screening are more likely to be diagnosed with GDM (16.3% vs 8.3% in the meta-analysis)

      • Patient with one-step screening are more likely to be started on medications (7.1% vs 3.8%)

      • Patients undergoing one-step screening were more likely to have NICU admission (5.1% vs 4.5%)

      • Patients undergoing one-step screening were more likely to have babies experience hypoglycemia (9.3% vs 7.6%)

      • Rates of LGA  babies are similar between strategies (8.8% one step, 9.2% two step)

      • Rate of primary cesarean delivery was similar between groups (24.0% one step vs 24.7% two-step).

    • What can we conclude from this?

      • One-step testing seems to lead to increased resource utilization (more diagnoses, more folks on treatment, more NICU admissions)

      • One-step testing does not appear to differ from two-step testing for some maternal short-term outcomes (LGA, cesarean delivery rate) or fetal outcomes (did not cover above but shoulder dystocia, RDS, stillbirth, neonatal death were all similar between groups)

      • We don’t have any significant evidence about long-term outcomes for mother or fetus (i.e., later-in-life diabetes diagnoses, obesity rates in offspring, etc.)

        • An editorial about the meta-analysis makes the case that one-step testing might still be cost-effective if the increased resource utilization means fewer downstream consequences… remains to be seen and tough to study!

  • Early GDM screening: do we have anything new?

    • Since our last podcast, there have been two US-based RCTs about this (Roeder, Harper)

      • In the Roeder paper, patients with an A1c of > 5.7% or fasting glucose of > 92 were randomized to hyperglycemia therapy and nutritional counseling at the time of enrollment (early pregnancy) vs usual timing (3rd trimester)

        • The study was ended early due to poor enrollment, but:

          • Treatment in early pregnancy didn’t improve maternal or neonatal outcomes, including fat mass, weight percentile, macrosomia, or maternal weight gain.

          • Treatment also didn’t significantly reduce the diagnosis rate of GDM at a usual-timing screening test (14.2% early treatment vs 25.8% usual treatment, p=0.17)

      • In the Harper paper, obese patients (BMI > 30) were randomized to a traditional two-step test in early pregnancy (14-20 weeks) versus traditional timing.

        • Those who screened negative early were also re-tested at traditional timing.

        • Early screening did not reduce a composite perinatal outcome, nor did it seem to affect other important secondary outcomes.

      • What can we take away from these papers?

        • We still have a ways to go on proving the value of the “early GDM screen,” particularly of doing multiple glucose challenge tests.

  • Guidance before the 1st step of the two-step approach:

    • One common patient question is whether fasting or eating anything in particular might make one more likely to “screen in” on the 50g, 1h OGTT.

      • In the January 2023 Green Journal, a group at Stanford randomized patients to a 6-hour fast prior to the 1hr test, versus eating within 2hr of the OGTT.

      • The “fed” group actually had a lower rate of screening positive (13%) versus the “fasting” group (31%).

        • Ultimately, the incidence of GDM was also higher in this “fasting” group (12.4% vs 5.1%). 

      • The group theorizes this is due to a phenomenon previously called “starvation diabetes,” in which fasting leads to an increase in glucagon and decrease in insulin, thus making one transiently glucose intolerant; and then later, insulin kicks back in and returns you to a normal metabolic state. 

      • This study only had about 100 individuals per arm of the trial, so hard to draw conclusions about neonatal/obstetrical outcomes, but none different in what they were able to assess.

    • What can we take away?

      • Hard to know totally, but probably don’t encourage fasting prior to the 1hr OGTT!

Gestational Diabetes Trio, Featuring A Special Interview with Dr. Donald Coustan

/

Happy Holidays to all, and to celebrate the season we have a very sweet triple episode release today! The first two episodes are focused on the pathophysiology, diagnosis, and treatment of GDM, while the third is a special interview with Dr. Donald Coustan, Professor and Chair Emeritus of the Department of Obstetrics and Gynecology at Brown University. Dr. Coustan was recently profiled by AJOG as a “Giant in Obstetrics and Gynecology.” We hope you enjoy the interview and his perspective on GDM and OB-GYN more generally.

The ACOG PB (PB 190) on GDM was recently updated in February 2018. There is also a new bulletin on Pregestational Diabetes (PB 201), though we don’t spend much time on pregestational diabetes today.

We discuss multiple ways to diagnose GDM, based on different organization’s recommendations. The classic Carpenter-Coustan criteria endorsed by ACOG and the National Diabetes Data Group (NDDG) are based on two-step testing. An initial 50 gram glucose tolerance test is performed, and patients move on to the second screen if their 1hr glucose is measured at 130-140 mg/dL, pending on the institution. It is generally accepted that a value >200 mg/dL is diagnostic without moving on to the second step.

The three hour test is based on a 100g glucose load. The cutoffs vary by time point. Two elevated values are needed to diagnose GDM; however, there is increased risk for the patient even with just one elevated value on three hour testing. The classic Carpenter-Coustan criteria as well as the NDDG criteria are shown here from PB 190:

ACOG PB 190: The Carpenter-Coustan criteria are the most commonly used in the USA.

There is also single-step testing proposed by the International Association for the Study of Diabetes in Pregnancy, that uses a 75g, two-hour glucose tolerance test. Any one elevated value (fasting > 92, 1 hour > 180, or 2 hour > 153) is diagnostic of GDM, and no second screen is needed. The ADA has endorsed these criteria recently but also admits that there is not clear-cut evidence to support one screening strategy over another. ACOG endorses the two-step screening at this time.

Much of the research regarding treatment of GDM that we review in the podcast is well-reviewed in PB 190, so we won’t rehash it here. If non-pharmacologic treatments fail (monitored fasting and postprandial blood glucose levels are consistently elevated), an oral agent or insulin is required, with insulin being the gold-standard. How do you initiate insulin? See our guide below!

And remember — postpartum patients with GDM need a 2 hour, 75 gram glucose tolerance test between 4 and 12 weeks postpartum to rule out type 2 diabetes. A fasting > 125 or a 2 hour > 200 is diagnostic. A fasting between 100-125 or a 2 hour between 140-199 demonstrates impaired glucose tolerance. And even with normal values, anyone with GDM has a 15-70% chance of developing T2DM later in life, so it’s an important part of the pregnancy history to correspond back to the patient’s PCP.