Advertisement
Letter to the Editor| Volume 60, ISSUE 3, P318-319, March 2007

Unexplained residuals models are not solutions to statistical modeling of the fetal origins hypothesis

      We read with great interest an article by Keijzer-Veen et al. [
      • Keijzer-Veen M.G.
      • Euser A.M.
      • van Montfoort N.
      • Dekker F.W.
      • Vandenbroucke J.P.
      • van Houwelingen H.C.
      A regression model with unexplained residuals was preferred in the analysis of the fetal origins of adult diseases hypothesis.
      ] in which they propose a regression model to overcome the problem of adjustment for current body weight in the fetal origins of adult disease hypothesis. Instead of adjusting for current body weight, Keijzer-Veen et al. proposed that the unexplained residual current body weight (CWres) should be adjusted for. The variable CWres comprises residuals of the simple regression model where current body weight is regressed on birth weight; Keijzer-Veen et al. considered CWres to be the difference between attained (current body weight) and predicted (by birth weight) adult body weight. They recognized that adjustment for current body weight might not be justified because current body weight falls on the pathway between birth weight and disease outcome, and they therefore prefer their alternative approach because it estimates the effects of birth weight and the effect of additional weight gain in a single model.
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Journal of Clinical Epidemiology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Keijzer-Veen M.G.
        • Euser A.M.
        • van Montfoort N.
        • Dekker F.W.
        • Vandenbroucke J.P.
        • van Houwelingen H.C.
        A regression model with unexplained residuals was preferred in the analysis of the fetal origins of adult diseases hypothesis.
        J Clin Epidemiol. 2005; 58: 1320-1324
        • Tu Y.-K.
        • West R.M.
        • Ellison G.T.H.
        • Gilthorpe M.S.
        Why evidence of fetal origins of adult diseases can be statistical artifact: the reversal paradox examined for hypertension.
        Am J Epidemiol. 2005; 161: 27-32
        • Lucas A.
        • Fewtrell M.S.
        • Cole T.J.
        Fetal origins of adult disease-the hypothesis revisited.
        BMJ. 1999; 319: 245-249
        • Wickens T.D.
        The geometry of multivariate statistic.
        Lawrence Erlbaum Associates, Hillsdale, NJ1995
        • Tu Y.K.
        • Ellison G.T.H.
        • Gilthorpe M.S.
        Revisiting four-model principle in fetal origins hypothesis.
        Pediatr Res. 2005; 58 ([abstract]): 1013