The prevalence of gestational diabetes is estimated at between 2 and 6%, but can be much higher in specific populations.

The specific treatment of gestational diabetes (diet, control weight gain, self monitoring glucose , insulin therapy) reduces complications severe perinatal, fetal macrosomia, and preeclampsia compared with abstention therapy, without additional risk of caesarean section.

Several early studies have shown that the determination of fructosamine is a very bad way of diagnosis for Gestational Diabetes. However, few studies have investigated the relationship between the determination of serum fructosamine and put under insulin in case of Gestational Diabetes.

The purpose of this study is to determine if serum fructosamines rate can be a predictive marker of the starting insulin at the patients suffering from Gestational Diabetes

Inclusion Criteria:

• Age ≥18 years
• Ongoing pregnancy with a start date of pregnancy known and confirmed by an ultrasound of the 1st Quarter performed between 10 and 14 weeks with measurement of the crown-rump length
• Hospitalization for a GD's discovery whatever the term of pregnancy (GD is defined with nationals recommendations of CNGOF)
• Not opposed to participate in the study

Exclusion Criteria:

• Pregnancy uncertain term (absence of early ultrasound)
• Previous diabetes to the pregnancy (type 1 or 2)
• Multiple Pregnancy
• No affiliation to a social security scheme
• Woman with a measure of legal protection