Docstoc

Maternal Alcohol Consumption and Risk of Orofacial Clefts

Document Sample
Maternal Alcohol Consumption and Risk of Orofacial Clefts Powered By Docstoc
					Maternal Alcohol Consumption
and Risk of Orofacial Clefts

           Lixian Sun
    Department of Epidemiology
    Orofacial Clefts

n   Birth prevalence of 1-2/ 1,000 births

n   Defect Type

    v Cleft lip only (CL)
    v Cleft palate only (CP)

    v Cleft lip with cleft palate (CLP)
      Alcohol Consumption

n   2002 Behavioral Risk Factor Surveillance
    System found over 50% of women of
    childbearing age (18-44) reported any alcohol
    use in previous 30 days of their pregnancies.
n   Among women who might become pregnant,
    55% reported any drinking, 13% reported drink
    more than 7 drinks per week and 12.4%
    reported binge drinking (=> 5 at one time)
    NBDPS

n   Multi-center (10 centers) study headed
    by CDC.

n   Identify risk factors for infants with birth
    defects.

n   Over 30 major birth defects were
    classified
  Hypothesis

H0: Periconceptional* alcohol
    consumption can increase the
    risk of having oralfacial clefts.

* periconceptional: from one month before
  pregnancy to the end of first trimester of the
  pregnancy
Case/Control Selection

n   Cases were identified from live births,
    fetal deaths and elective terminations
n   Controls included only live births
    without any major birth defect.
n   All case and control moms completed
    telephone interview.
Alcohol Exposure

v   No alcohol consumption
    (reference group)

v   Any alcohol consumption
Sample Size

 Cases and Controls using in this
 project were born from Jan 1 to
 Dec 31, year 2005

        Cases:    286
        Controls: 789
     Methods

n   Both non-informative priors and
    informative priors were used
n   GLM with logit link function was used to
    estimate the odds ratio
n   Convergence was checked
n   Potential confounders (Baby’s gender and
    Maternal cigarette smoking from B1 to
    P3) were adjusted
Using Non-informative Priors


    Alpha ~ dnorm (0, 0.01)
    Betas ~ dnorm (10,0.01)
Running 1000 iterations
   Auto correlation
History plots
BGR Diagnostic
Re-run 19000 more iterations
and burn in 10000
       Summary of Statistics


node   mean     sd       2.5%    97.5%

 OR    0.8631   0.1293   0.854   1.139
Using Informative Priors


 beta1~dnorm(0.0431, 1312)
 beta2~dnorm(0.3135, 349)
 beta3~dnorm(0.3593, 249)
 alpha ~dnorm(-1.298, 528)
Run 1000 iterations
 Auto correlation
History plot
BGR diagnostics
       Re-run 3000 iterations and
         burn in 1000 iterations
         Summary of Statistics

node     mean   sd       2.5%     97.5%

OR      1.038   0.0282   0.9837   1.096
    Summary

n No evidence to show that
  alcohol is an effect factor of
  oral facial cleft.
n Using informative priors

  provides narrower credible set
n Less iterations were needed if

  informative prior was used.
           More analyses

n Examine the effect of alcohol
  consumption and different
  phenotype of oralfacial clefts.
n Examine the effect of alcohol

  by alcohol type.
n Examine the effect of alcohol

  by exposure of binge drinking.
 Acknowledgements

•All participated families

•Professor Cowles

•Every one in this class

•My daughter for her coming back

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:7/14/2013
language:English
pages:24