SWIFT: Study of Women, Infant Feeding and Type 2 Diabetes After GDM Pregnancy

The SWIFT study prospectively enrolled 1,035 women diagnosed with GDM during pregnancy who
delivered a singleton, live born infant of at least 35 weeks gestation at a Kaiser Permanente
Northern California hospital from 2008 to 2011, and met other eligibility criteria. Women
with recent GDM were enrolled into the study at 6-9 weeks postpartum (study baseline) and
consented to annual follow up exams during the subsequent two years. Study enrollment began
in late 2008 and ended in December 2011. The SWIFT cohort is racially and ethnically diverse
(75% minority) with 35% Asian, 9% Black, 31% Hispanic, 23% White, and 2% other race groups.

Women provided written informed consent for three in-person exams involving administration of
the 2-hour 75 gram Oral Glucose Tolerance Test (OGTT) and numerous other assessments at each
exam to classify glucose tolerance in women following GDM pregnancy. At baseline, 21 women
classified with diabetes were excluded from follow up, 2 dropped out, and 2 were ineligible.
There were 1,010 women with recent GDM and without diabetes at baseline followed to evaluate
the primary study outcome, the progression to glucose intolerance during the two year study
follow up period, defined as incident diabetes by American Diabetes Association (ADA)
criteria from the 2-hour 75 gram OGTT glucose values, and/or clinical medical diagnosis of
diabetes. The annual study follow up exams during the 2 years post-baseline occurred from
2009-2014. The study has continued ongoing cohort surveillance for progression to diabetes
and other subsequent pregnancies for participants via the clinical diagnoses within the
Kaiser Permanente electronic health records.

The primary exposure for the study is lactation intensity and duration assessed
quantitatively using the method by Piper et al. 2001 to estimate a continuous lactation
intensity and duration score up to 12 months postpartum. The study assessed infant feeding
prospectively from prenatal telephone contacts to assess breastfeeding intention via a
standardized method, in-patient hospital delivery records, participant infant feeding
diaries, telephone contacts at 1 month postpartum, self-administered monthly mailed surveys
(from 3 to 11 months postpartum), and from surveys at the three in person annual study exams.
Covariable data collection during and after pregnancy was obtained from electronic medical
records related to perinatal course (e.g., laboratory diagnosis of GDM severity: 3-hr OGTT
z-score and GDM treatment, gestational age at GDM diagnosis, maternal pre-pregnancy BMI,
gestational weight gain), medical history, prenatal measures, subsequent pregnancies, and the
maternal and newborn outcomes. At each of three study exams, trained research staff assessed
maternal characteristics (sociodemographics, medical and reproductive history, subsequent
pregnancies, medication use, recurrence of GDM, physical activity, dietary intake,
depression, and sleep habits/disorders), as well as breastfeeding intensity and duration,
infant health, and complementary infant dietary intake using self- and
interviewer-administered questionnaires. Trained research staff measured participant
anthropometry, skinfold thickness, and body composition via bioelectrical impedance according
to standardized research protocols, collected, processed and stored biospecimens (blood and
buffy coat), and administered questionnaires at each in-person exam. The SWIFT Offspring
Study, an ancillary study to SWIFT of the mother-infant dyads, conducted three in-person
exams to evaluate infant growth during the first year of life as well as sleep, temperament,
and dietary intake, and collect saliva specimens in the infants of the SWIFT mothers.

Ongoing surveillance to ascertain new diagnoses of diabetes in the SWIFT cohort occurred both
during and after the study period via the KPNC electronic health records from 2009 to 2018.
The SWIFT study plans to recontact all 1,033 active participants in 2019 to conduct a fourth
in-person exam at 10 years post-baseline from 2019-2022. The fourth exam at 10-years
post-baseline will reassess glucose tolerance, anthropometry, body composition and other
attributes as described previously. The SWIFT study has also utilizes fasting plasma
specimens sampled within the early postpartum period to identify metabolites for the early
prediction of future progression to type 2 diabetes. Changes in anthropometry, lifestyle
behaviors, and risk factors for diabetes will also be assessed at the 10-year post-baseline
in-person exam.

Inclusion Criteria:

- age 20 to 45 years at delivery

- received prenatal care in Kaiser Permanente Northern California (KPNC) health care
system

- GDM pregnancy diagnosed using the 3-hour 100 g OGTT by Carpenter and Coustan criteria

- delivered a singleton, live birth >= 35 weeks gestation

- no pre-existing diabetes or other serious medical conditions prior to index GDM
pregnancy

- no diabetes diagnosis (2-hour 75 gram OGTT) at 6 to 9 weeks postpartum for the index
GDM pregnancy

- no use of thyroid medications, steroids, or other medications affecting glucose
tolerance

- not planning to move from the northern California area within the subsequent 24 months

- not planning another pregnancy within the next two years

- Two infant feeding groups: women who did not lactate or did so for less than 3 weeks,
OR women who provided no supplemental milk feeds at 2-4 weeks and planned to continue
intensive lactation defined as <= 1 formula supplement (6 oz/day) from 6-9 weeks until
4 months or more postpartum.

Exclusion criteria:

- women who fed breast milk and 7-16 oz of formula (mixed feeding) during the first 4
weeks of life