Title:
Effect of smoking on the development of gestational diabetes

Authors:
Timothy N Dunn MD, Yumo Xue MS, Ruby Knupp, MD, Brian M. Casey, MD, Alan N. Tita, MD, PhD, Lorie M. Harper MD MCSI

Background:

• Smoking has been associated with hyperinsulinemia and insulin resistance, and it treated as a causal risk factor for type 2 diabetes in non-pregnant adults.
• While ACOG and the ADA recommend screening patients with risk factors for pregestational diabetes or early-onset gestational diabetes (GDM) at the first prenatal visit, smoking is not listed as a risk factor.
• Current literature regarding GDM and smoking is sparse, with a mixed picture of the relationship of smoking as a risk factor for GDM.

Specific Aims:

• To determine if smoking is associated with an increased risk for GDM

Methods:

• Unplanned secondary analysis of a multi-center, randomized controlled trial that compared early to routine screening for GDM in women with a BMI ≥30kg/m².
• Patients excluded from original trial if they had a prior cesarean delivery, pre-existing diabetes mellitus, history of bariatric surgery, major medical illnesses (such as cardiac disease or sickle cell disease among others)
• Patients included in this secondary analysis if smoking history was recorded at time of enrollment
• Primary analysis performed using dichotomous exposure of self-reported smoking, with patients reporting “yes” or “no.”
• Patients who reported smoking were stratified to those who smoked ≥10 cigarettes per day, < 10 cigarettes per day, or reported smoking formerly.
• Primary outcome was development of GDM between the exposed (smoking) and unexposed (nonsmoking) cohorts.
• Secondary outcomes included the type of GDM (A1 or A2), 50g 1-hour glucose challenge results, HbA1c levels and perinatal outcomes. Perinatal outcomes were a composite of macrosomia, primary cesarean delivery, pregnancy-induced hypertension (gestational hypertension, pre-eclampsia, or eclampsia), shoulder dystocia, neonatal hypoglycemia, or neonatal hyperbilirubinemia.
• For statistical analysis, groups were composed using chi-squared or student’s t-test for trend as appropriate. Multivariable logistic regression was used to adjust for baseline differences between groups.

Results:

• 221 (24%) of the 913 patients reported smoking.
• Smokers were less likely to be Hispanic or married, and they were more likely to report substance or alcohol use or to have chronic hypertension
• GDM incidence was similar between smoking and non-smoking groups
• After controlling for differences between groups, GDM was still not increased by smoking.
• One-hour GCT at 24 weeks was lower in smokers than nonsmokers, but both values remained below 135mg/dL.
• When stratifying groups by smoking amounts, GDM and secondary outcomes remained similar between groups.

Table 1: Primary and Secondary Outcomes

Table 2: Outcomes stratified by smoking amount

Figure 1: Gestational Diabetes Incidence per Smoking Amount

Conclusions:

• In an obese patient population, smoking was not related to GDM development.
• There was no dose-response link between smoking and GDM.