Obesity, Pregnancy Weight Gain and Preeclampsia as Risk Factors for Preterm Birth: Evidence from a D

1
Obesity, Pregnancy Weight Gain and Preeclampsia
as Risk Factors for Preterm Birth Evidence from
a Diverse, Urban Cohort
Terry J. Rosenberg, PhDSamantha Garbers,
MPAMary Ann Chiasson, DrPH2005 AMCHP ANNUAL
CONFERENCE
2
Overview

• Goal Examine associations of obesity, excess
  prenatal weight gain, preeclampsia, and preterm
  birth
• Data New York City Births File (birth
  certificate data) 330,216 singleton births in
  1999-2001 with mothers weight data
• Methods Chi-square ANOVA logistic regressions
  to predict preeclampsia and to predict preterm
  birth

3
Definitions

• Obesity - Prepregnancy weight of 200 or more
  pounds (gt91 kg)
• Excess prenatal weight gain 41 or more pounds
  (gt18 kg) during pregnancy
• Preeclampsia - As diagnosed by physician
  completing birth certificate standard ACOG
  definition
• Preterm birth Before 34 weeks of gestation

BMIs could not be computed because there are no
height data on the birth certificate.
4
Key Variables in Analysis

• Obesity 6.3 (n20,702)
• Excess weight gain 18.4 (n60,695)
• Preeclampsia 2.1 (n7,011)
• Preterm birth 2.1 (n7,011)

5
Sample CharacteristicsRace / Ethnicity
6
Sample Characteristics
Missing data excluded for each variable.
7
Sample Characteristics
Missing data excluded for each variable.
8
Percent Obese by Race / Ethnicity
Obese 200 or more pounds prepregnancy Signific
ance level for Chi-square plt.001
9
Excess Prenatal Weight Gain by Race / Ethnicity
Excess prenatal weight gain 41 or more
pounds. Significance level for Chi-Square plt.001
10
Percent Preeclampsia by Race / Ethnicity
Significance level for Chi-square plt.001
11
Percent Preterm Birth by Race / Ethnicity
Significance level for Chi-square plt.001
12
Logistic Regressions to Predict Preeclampsia and
Preterm Birth Confounders Included in All Models

• Age
• Race/ethnicity
• Marital status
• Educational attainment
• Parity (no prior births, 1 prior, 2 prior, 3 or
  more prior)
• Timing of initiation of prenatal care
• Insurance coverage (Medicaid as low-income)
• Chronic and pregnancy-related diabetes
• Chronic and pregnancy-related high blood
  pressure
• Preeclampsia

In preeclampsia regression only In
preterm birth regression only
13
Logistic Regressions to Predict Preeclampsia and
Preterm BirthWeight Variables Included in All
Models

• Prepregnancy Weight (pounds)
• Under 100
• 100-149 (reference group)
• 150-199
• 200-299 (obese)
• 300 or more (obese)
• Prenatal Weight Gain (pounds)
• 0-40 (reference group)
• 41 or more (excess)

14
Part 1 of Logistic Regression to Predict
PreeclampsiaSignificant Confounders
15
Part 2 of Logistic Regression to Predict
PreeclampsiaAORs for Weight Variables
The reference group is 100-149 pounds. The
reference group is 0-40 pounds.
16
Part 1 of Logistic Regression to Predict Preterm
BirthSignificant Confounders
17
Part 2 of Logistic Regression to Predict Preterm
BirthAORs for Weight Variables
The reference group is 100-149 pounds. The
reference group is 0-40 pounds.
18
Excess Weight Gain by Prepregnancy Weight
Significance level for Chi-square plt.001
19
Incremental Mean Prepregnancy Weight with
Increasing Parity
Significance level for ANOVA p lt.001
20
Conclusions

• Obesity and excess prenatal weight gain increase
  the risk of preeclampsia for all women
• The greatest increased risk of preeclampsia is
  among women weighing 300 pounds or more (136kg or
  more) prepregnancy
• Preeclampsia is the most significant predictor of
  preterm birth
• The negative effects of obesity and excess
  prenatal weight gain on preterm birth are
  mediated through their effects on preeclampsia

21
Recommendations

• Nutrition and exercise counseling are needed
  prepregnancy and during the interpartum interval
• Moderate weight gain should be encouraged during
  pregnancy
• Programs should address prevention of incremental
  weight increase with each pregnancy

22
Contact Information

• trosenberg_at_mhra.org