The n e w e ng l a n d j o u r na l of m e dic i n e

4. Cole TJ, Faith MS, Pietrobelli A, Heo M. What is the best findings of Flegal et al. that many overweight
measure of adiposity change in growing children: BMI, BMI %,
BMI z-score or BMI centile? Eur J Clin Nutr 2005;59:419-25. [Er-
children may not, in fact, have excess body fat.3
ratum, Eur J Clin Nutr 2005;59:807.] Unfortunately, the assessment of body composi-
tion was not feasible in a school setting and in a
sample of more than 6000 children. We do note
The author replies: My colleagues and I agree that the percentage of children with a waist cir-
with Bachman about the need to approach obe- cumference (a measure that can serve as an esti-
sity prevention from both sides of the energy bal- mate of truncal fat) greater than the 90th per-
ance equation. Our HEALTHY intervention includ- centile was significantly less in the intervention
ed, as one of its four intervention components, a schools than in the control schools. These data,
curriculum to increase the quantity and quality along with the greater reductions in the preva-
of school-based physical education.1 In addition, lence of obesity in the intervention schools, sug-
much of our classroom-based program and com- gest that our intervention had effects on obese
munications strategies focused on increasing phys- children, a group shown by Flegal et al. to have
ical activity and decreasing sedentary behavior.2 excess body fat.3
We agree with Singhal and Misra that the field Gary D. Foster, Ph.D.
lacks consensus about the single best measure to Temple University
assess changes in adiposity among growing chil- Philadelphia, PA
dren. Accordingly, although our primary outcome gfoster@temple.edu
was categorical and based on percentile of body- for the HEALTHY Study Group
mass index (the combined prevalence of over- Since publication of his article, the author reports no further
weight and obesity or a body-mass index >85th potential conflict of interest.
percentile), we included multiple categorical and
1. The HEALTHY Study. Intervention. (http://www.healthystudy
continuous secondary measures (the prevalence .org/intervention.htm.)
of obesity, z score for body-mass index, and waist 2. Idem. Materials matrix. (http://www.healthystudy.org/
circumference) to broadly assess changes in adi- materialsmatrix.htm.)
3. Flegal KM, Ogden CL, Yanovski JA, et al. High adiposity and
posity. We also agree that it would have been high body mass index-for-age in US children and adolescents
optimal to assess body composition, given the overall by race-ethnic group. Am J Clin Nutr 2010;91:1020-6.

Molecular Architecture of the Goodpasture Autoantigen

To the Editor: Pedchenko et al. (July 22 issue)1
report the specificity and molecular architecture
of epitopes that bind autoantibodies in patients
with Goodpasture’s disease and show that per-
turbation of the quaternary structure of the
α345NC1 hexamer plays an important role in the
development of this disorder. Goodpasture’s dis-
ease is an organ-specific autoimmune disorder
characterized by rapidly progressive glomerulo-
nephritis and pulmonary hemorrhage, with lin-
ear deposits of antibodies along the glomerular
basement membrane (GBM). Therefore, it is of
interest to determine whether there are any dif-
ferences in the epitope specificity and autoanti-
body titer associated with Goodpasture’s disease
on the basis of whether the disease is evident in
lung tissue. A study characterizing the antigenic
epitopes of the noncollagenous-1 (NC1) domain
and positive results for antineutrophil cytoplas-
mic antibody (ANCA) would be informative, since
41% of the patients with anti-GBM antibodies in
this report were also positive for ANCA and since
ANCA and anti-GBM antibodies are reported to
be positive in some patients with crescentic glo-
merulonephritis.2 Such a study would help draw a
clearer picture of what Pedchenko et al. call “con-
formeropathy” in autoimmune vasculopathy.
Hiroshi Okamoto, M.D., Ph.D.
Minami–Otsuka Clinic
Tokyo, Japan
No potential conflict of interest relevant to this letter was re-
ported.
1. Pedchenko V, Bondar O, Fogo AB, et al. Molecular architec-
ture of the Goodpasture autoantigen in anti-GBM nephritis.
N Engl J Med 2010;363:343-54.
2. Rutgers A, Slot M, van Paassen P, van Breda Vriesman P,
Heeringa P, Tervaert JW. Coexistence of anti-glomerular basement
membrane antibodies and myeloperoxidase-ANCAs in crescen-
tic glomerulonephritis. Am J Kidney Dis 2005;46:253-62.

1770 n engl j med 363;18  nejm.org  october 28, 2010

The New England Journal of Medicine

Downloaded from www.nejm.org at Hinari Phase 1 sites -- comp on November 25, 2010. For personal use only. No other uses without permission.
Copyright © 2010 Massachusetts Medical Society. All rights reserved.
 correspondence
The authors reply: Okamoto suggests that de-
termining whether there are differences in titers
of autoantibodies and epitope specificity in
Goodpasture’s disease on the basis of lung in-
volvement might be beneficial for understanding
the mechanism of this disease. In one study, pa-
tients with Goodpasture’s disease who had posi-
tive results for both ANCA and anti-GBM anti-
bodies had a poor prognosis when they presented
with severe renal failure.1 Other studies reported
no difference between patients with ANCA and
those without ANCA with respect to the NC1-
antigen specificity of anti-GBM antibodies.2 In
our study, 12 of 46 patients had lung hemor-
rhage. However, there was no difference in auto-
antibody titers between patients with hemorrhage
and those without hemorrhage. Furthermore, de-
spite the earlier suggestion that the prognosis for
double-positive patients may be dependent on
both populations of antibodies,3 we did not find
any difference between patients with ANCA and
those without ANCA in titers of circulating anti-
α3NC1 or anti-α5NC1 autoantibodies. Our analy-
Valproic Acid Use in Pregnancy and Congenital Malformations
To the Editor: In their article describing val-
proic acid monotherapy in pregnancy and major
congenital malformations (June 10 issue),1 Jen-
tink and colleagues concluded by recommending
that the use of valproic acid be avoided, if possi-
ble, in women of childbearing potential. However,
this study did not assess the effect of the dose of
valproic acid on the risk of birth defects. In many
patients with primary generalized epilepsy, the
condition is well controlled with the use of low
doses of valproic acid (<1000 mg). Among preg-
nant women in the Australian Pregnancy Register,
the risk of teratogenic effects was not significant-
ly higher with valproic acid than with alternative
antiepileptic drugs, and seizure control was better
in patients treated with valproic acid than in those
treated with lamotrigine or carbamazepine, in any
trimester of pregnancy.2 Seizures endanger the
mother as well as the fetus. We believe that a
more appropriate conclusion is that high doses
of valproic acid should not be used as a first-
choice drug in women of childbearing potential.
n engl j med 363;18  nejm.org  october 28, 2010
The New England Journal of Medicine
Downloaded from www.nejm.org at Hinari Phase 1 sites -- comp on November 25, 2010. For personal use only. No other uses without permission.
Copyright © 2010 Massachusetts Medical Society. All rights reserved.
sis of antibody epitopes indicates that conforma-
tional changes occur in collagen IV α345NC1
with no difference in the presence of concomi-
tant ANCA. Thus, any influence of ANCA on the
severity of lesions among patients with Goodpas-
ture’s disease does not appear to be mediated by
changes in epitope conformation.
Agnes B. Fogo, M.D.
Billy G. Hudson, Ph.D.
Vadim Pedchenko, Ph.D.
Vanderbilt University Medical Center
Nashville, TN
billy.hudson@vanderbilt.edu
Since publication of their article, the authors report no fur-
ther potential conflict of interest.
1. Levy JB, Hammad T, Coulthart A, Dougan T, Pusey CD. Clin-
ical features and outcome of patients with both ANCA and anti-
GBM antibodies. Kidney Int 2004;66:1535-40.
2. Hellmark T, Niles JL, Collins AB, McCluskey RT, Brunmark
C. Comparison of anti-GBM antibodies in sera with or without
ANCA. J Am Soc Nephrol 1997;8:376-85.
3. Short AK, Esnault VL, Lockwood CM. Anti-neutrophil cyto-
plasm antibodies and anti-glomerular basement membrane an-
tibodies: two coexisting distinct autoreactivities detectable in
patients with rapidly progressive glomerulonephritis. Am J Kid-
ney Dis 1995;26:439-45.
Frank Vajda, M.D.
University of Melbourne
Parkville, VIC, Australia
vajda@netspace.net.au
Terence O’Brien, M.D.
Royal Melbourne Hospital
Parkville, VIC, Australia
Drs. Vajda and O’Brien report being investigators for the Aus-
tralian Pregnancy Register, which has received grant support
from UCB Pharma, Parke-Davis (now Pfizer), Novartis, Sanofi-
Aventis, Novo Nordisk, Janssen Cilag, Roche, Biogen, Glaxo­
Smith­K line, CSL Behring, Schering, and Aspen. No other poten-
tial conflict of interest relevant to this letter was reported.
1. Jentink J, Loane MA, Dolk H, et al. Valproic acid monother-
apy in pregnancy and major congenital malformations. N Engl J
Med 2010;362:2185-93.
2. Vajda FJ, Hitchcock A, Graham J, et al. Foetal malformations
and seizure control: 52 months data of the Australian Pregnancy
Register. Eur J Neurol 2006;13:645-54.
The Authors and a Colleague Reply: As we
mentioned in our article, we did not have infor-
mation on doses of valproic acid; therefore we
cannot draw any conclusion about doses of val-
proic acid and the risk of congenital malforma-
1771