Content uploaded by Lucia Billeri
Author content
All content in this area was uploaded by Lucia Billeri on Oct 02, 2014
Content may be subject to copyright.
Clin Chem Lab Med 2008;46(5):687–690 2008 by Walter de Gruyter •Berlin •New York. DOI 10.1515/CCLM.2008.131 2007/485
Article in press - uncorrected proof
Short Communication
Time to reconsider the clinical value of
immunoglobulin G4 to foods?
Daniela Bernardi, Franco Borghesan, Diego
Faggian, Fulvia Chieco Bianchi, Elisabetta
Favero, Lucia Billeri and Mario Plebani*
Department of Laboratory Medicine, University
Hospital of Padova, Padova, Italy
Abstract
Background: The usefulness of serum antibodies to
common food antigens (immunoglobulin G4; IgG4)
assay in management of patients suffering from food
intolerance was assessed.
Methods: A total of 22 asymptomatic healthy subjects
and 68 patients with symptoms referred for suspected
food intolerance were studied. Serum IgG4 to 19
common foods was measured by an automated
immunoassay.
Results: The area under the receiver operating char-
acteristic curve was 0.92 (standard error 0.04) and, at
a threshold value of 2.3 U/mL, the IgG4 determination
had a sensitivity of 0.81, with a specificity of 0.87.
With a pre-test probability of 5% and 20%, the post-
test probability of having disease was found to be
24% and 61%, respectively, and 1.1% and 5% if the
result was negative. Cohen’s kvalue (0.83) indicated
a good agreement between symptoms and IgG4
concentrations.
Conclusion: Serum IgG4 assay may play a role in rul-
ing out food intolerance, because of its satisfactory
negative predictive value (0.99).
Clin Chem Lab Med 2008;46:687–90.
Keywords: diagnostic accuracy; exclusion diet; food
intolerance; food-specific IgG4 antibody.
Food intolerance is recognised in up to one-fifth of the
general population (1). Currently, the management of
this condition consists of exclusion and reintroduction
diets, with the subsequent confirmation of outcomes
by means of the double-blind vs. placebo test. How-
ever, an approach involving the identification of the
offending food(s) by dietary elimination and re-chal-
lenge can be cumbersome, and poor patient compli-
ance can compromise its clinical efficacy. Because it
is difficult to diagnose food intolerance, its true prev-
*Corresponding author: Dr. M. Plebani, Department of
Laboratory Medicine, University Hospital of Padova,
Via Giustiniani 2, 35128 Padova, Italy
Phone: q39-0498212792, Fax: q39-049663240,
E-mail: mario.plebani@unipd.it
Received October 24, 2007; accepted January 28, 2008;
previously published online February 26, 2008
alence is unknown, but it has been estimated to occur
in approximately 5% of the general population (2–4).
Moreover, our understanding of the pathophysiology
of food intolerance is incomplete, and this drawback
is paralleled by a paucity of options available for the
diagnostic work-up. Elevated values of serum IgG4
(immunoglobulin G4 subclass) antibodies to specific
food antigens, before dietary exclusion, may prove
useful in targeted dietary exclusion, obviating the
need to exclude a large number of foods from the
diet. We therefore investigated the appropriateness of
using in vitro diagnostics for food intolerance based
on food IgG4 determination in order to evaluate the
potential role of this measurement method in patient
management.
Serum IgG4 concentrations were evaluated in sub-
jects classified with no symptoms associated with
food ingestion and following a free diet without any
restrictions (healthy controls). To assess the IgG4 con-
centrations in food intolerance, sera of patients clas-
sified with adverse food reactions were determined
(study group). Although perceived food intolerance
was not verified by a double-blind placebo controlled
food challenge, subjects received a food-specific IgG4
antibody-guided exclusion diet for at least 2 months.
At the end of the regimen period, all patients reported
their symptoms and, if feasible, were given a repeat
IgG4 assay.
Subjects in the control group, consisting of 22
asymptomatic healthy subjects (median age 39 years,
range 25–55 years; 4 males and 18 females) recruited
from the hospital staff, were asked specific questions
about the absence of bowel symptoms, atopic der-
matitis, bronchial asthma, headache related to food
ingestion, pruritus without dermatitis, gastroenteritis
and antibiotic consumption in the month prior to the
start of the study, because transient alterations in gut
permeability in healthy individuals, caused by various
factors (gastroenteritis, antibiotics, altered microbial
flora, stress), unlike in patients, may induce a tran-
sient IgG4 response to food antigens (5).
The study group consisted of 68 consecutive
patients (median age 36 years, range 10–71 years; 26
males and 42 females) who were referred to our clin-
ical service from September 2003 to January 2005
with symptoms, such as meteorism, diarrhoea, func-
tional dyspepsia, food intake related headache. Clini-
cal evaluation ruled out adverse reaction to lactose
(evaluated by lactose breath test) or celiac disease
(evaluated by anti-transglutaminase antibody), patho-
logical gastrointestinal diseases and psychological
disorders presenting with gastrointestinal symptoms.
688 Bernardi et al.: Time to reconsider the clinical value of IgG4 to foods?
Article in press - uncorrected proof
Figure 2 ROC curves for IgG4, in bold font and broken line,
when comparing controls with patients whose symptoms
resolved following milk or egg white exclusion diet,
respectively.
Figure 1 Egg white and cow’s milk specific serum IgG4
mean values (U/mL) in control and patient groups.
Following the Helsinki II declaration, the design and
execution of the experiment was thoroughly
explained to the subjects, and informed consent was
obtained from all subjects. The patients received a
food-specific IgG4 antibody-guided exclusion diet for
at least 2 months. In particular, they received a list of
the foods they had been advised to eliminate and tele-
phone contact details where they could contact per-
sonnel for further advice if necessary. Because 17
subjects either failed to follow the dietary indications
or were not contactable, response to the regimen was
established in 51 patients. A total of 19 patients,
responding well to dietary exclusion after the first
assay, underwent a second IgG4 assay test after the
diet.
Serum IgG4 antibody concentrations to common
foods, including milk, egg white, wheat, casein, rice,
yeast, potatoes, peanuts, cod fish, chicken, lamb,
beef, pork, tomatoes, carrots, onions, apples, bananas
and soy beans, were measured. Blood samples were
left to stand for 20–30 min before being centrifuged
at 3000 cycles/min (1300=g) for 15 min. The serum
was separated and frozen at –208C for subsequent
analysis. Samples were processed using a commer-
cially available immunoassay (Enea Specific IgG4,
BioAllergy International, Trieste, Italy). The antibody
titers were expressed as U/mL and the measured
range was between 0 U/mL and 30 U/mL. Inter- and
intra-assay variation coefficients ranged from 35%
(mean 10.5 U/mL) to 23% (mean 27.00 U/mL), with
some further differences among allergens.
The mean difference between the two study groups
was analyzed using the unequal variance t-test. A
p-value -0.05 was considered statistically significant.
Receiver operating characteristic (ROC) curve anal-
ysis (6) defined the IgG4 cut-off with the best diag-
nostic sensitivity and specificity and the highest
diagnostic power to discriminate between patients
classified with adverse food reaction and control
subjects. In view of the difficulty in diagnosing food
intolerance, we defined as true positives the cases in
which patients reported a dramatic reduction in all the
symptoms over 2 months following the food-specific
IgG4 antibody-guided exclusion diet.
Yet, also in view of the known diagnostic difficul-
ties, the true prevalence of food intolerance remains
unknown. We therefore calculated the accuracy meas-
urements of the test in two different populations:
patients referred to general practitioners and those
referred to allergy specialists. In the former category,
the prevalence of subjects reporting a complaint fol-
lowing the ingestion of a particular food was esti-
mated at approximately 5%; in the second category,
the estimated prevalence increased to approximately
20%, because patients had already been screened for
food protein-induced enteropathy, such as celiac
disease, or for lactose deficiency.
Agreement between categorical data was measured
by Cohen’s k, the chance corrected proportional
agreement (7). Although no objective criteria were
available for judging intermediate values, kis often
considered to provide agreement, which is poor if
-0.2, fair if 0.21-k-0.40, moderate if 0.41-k-0.60,
substantial if 0.61-k-0.80 and good if k)0.80.
IgG4 concentrations to milk, egg white, wheat, rice,
pork, tomatoes, apples and bananas were significant-
ly (p-0.001) lower in the control subjects than in
patients classified with adverse food reaction. In par-
ticular, egg white and cow’s milk specific IgG4 (the
most frequently represented antibodies in our popu-
lation) were 0.8"0.3 and 3.1"2.1 U/mL wmean and
standard error (SE)xin the control group, respectively.
In the patient group, IgG4 values were 11.9"1.7 and
13.7"1.6 U/mL, respectively (Figure 1).
Figure 2 shows the results of the ROC curves anal-
ysis for the cut-off value. The area under the ROC
curve (AUC) of IgG4 specific for cow’s milk (Figure 2,
AUC in bold font) was 0.89 (SE 0.06) and, at a thres-
hold value of 2.8 U/mL, IgG4 determination had a
sensitivity of 0.83, with a specificity of 0.92. When
comparing controls with patients classified with
adverse reaction to egg white, the AUC (Figure 2, bro-
ken line) was 0.87 (SE 0.07), and at a threshold value
of 2.0 U/mL, IgG4 determination showed a sensitivity
of 0.73, with a specificity of 0.83.
When prevalence was considered 5% (current esti-
mate), the positive predictive value (PPV) was low
(0.35 for cow’s milk IgG4; 0.25 for egg white IgG4) and
the relative number of false positives high, although
both sensitivity and specificity were good. The nega-
tive predictive values (NPV) were 0.99. As expected,
on increasing pre-test probability value to 20%, PPV
Bernardi et al.: Time to reconsider the clinical value of IgG4 to foods? 689
Article in press - uncorrected proof
Figure 4 Pre- and post-test probability of IgG4 to egg white
determination.
Figure 3 Pre- and post-test probability of IgG4 to cow’s milk
determination.
also increased to 0.72 and 0.61, respectively, and NPV
maintained the value of 0.95.
The likelihood ratio was calculated to estimate the
extent to which the test result would change the odds
of having intolerance, and, for a positive value (LRq)
of IgG4 measurement it was 10.38 for cow’s milk and
6.23 for egg white, meaning that the odds of having
the syndrome increased ten- or six-fold, respectively,
when the test is positive.
The likelihood ratio for a negative value (LR–) in
IgG4 measurement was 0.18 for cow’s milk IgG4 and
0.22 for egg white, meaning that the odds of having
the syndrome decreased by 0.05 when the test is
negative.
When the diagnostic test result was positive, the
post-test probability was 35% (milk), 25% (egg white),
and 72% (milk) and 61% (egg white) with a pre-test
probability of 5% and 20%, respectively. With a neg-
ative result, the post-test probability was 1% (milk and
egg white) and 4%–5% (milk and egg white) with a
pre-test probability of 5% and 20%, respectively
(Figures 3 and 4).
Cohen’s kvalue, an index of agreement between
symptoms and IgG4 concentrations, was found to be
0.83.
The exclusion diet and the double-blind food chal-
lenges are currently considered the best available
methods of identifying food to which patients are
intolerant. However, a biomarker that could accurate-
ly diagnose symptomatic food intolerance would
greatly facilitate clinical practice. The results of our
study, showing a significant difference between the
control and patient groups suggest that IgG4 deter-
mination may play a role in differentiating subjects
with from those without food intolerance. In partic-
ular, the threshold value selected provides the best
diagnostic sensitivity and specificity for IgG4 deter-
mination, thus it appears to be an assay with a good
diagnostic accuracy. However, sensitivity and speci-
ficity do not solve the problems of the prevalence of
the condition in different populations worldwide.
Therefore, the IgG4 predictive value was calculated,
providing a low PPV in the group with a prevalence
of 5%. Thus, in screening the general population,
many individuals with false positive test results would
be obtained. On the other hand, the NPV was very
good, showing that the condition can be ruled out if
the screening result is negative. As expected, on
increasing pre-test probability value to 20%, the PPV
increased accordingly and the NPV value was main-
tained; this means that the test can be used to rule
out the condition. Moreover, we quantified the proba-
bility of a patient having the disease by considering
the post-test odds after measuring serum IgG4 and a
negative result appears to have a greater impact than
a positive test result, in particular with a prevalence
of 5%. This confirms that the test is better at ruling
out food intolerance than ruling it in, thus sparing the
patient an unnecessarily restrictive diet and allowing
appropriate treatment to be promptly initiated. Fur-
thermore, an evaluation was made of the agreement
between IgG4 concentrations and clinical data by
assessing the patient’s response (based on the
patients’ symptoms and IgG4 variation) to food-spe-
cific IgG4 antibody-guided exclusion diet over
2 months. Cohen’s kvalue indicated a good agree-
ment. The response after 2 months was, in fact,
690 Bernardi et al.: Time to reconsider the clinical value of IgG4 to foods?
Article in press - uncorrected proof
encouraging, with symptoms resolving in 78.5% of
subjects. The IgG4 results following the second assay,
available in 19 subjects, showed that IgG4 values
decreased after 2 months of diet in 89.5% of these
patients.
The present study has several limitations: a) the
number of control subjects was small; b) the value for
decreased IgG4 after the exclusion diet should have
been evaluated in a larger number of patients in a
prospective trial; and c) food responsible for symp-
toms was reintroduced in only a few cases for both
ethical and organizational reasons. However, our pre-
liminary data suggest that serum IgG4 could accu-
rately rule out symptomatic food intolerance and
would greatly facilitate clinical practice. However, the
non-satisfactory PPV does not allow clinicians to use
it as a definitive confirmatory tool. Further studies
should confirm these data in a more representative
number of patients and controls.
References
1. Young E, Stoneham MD, Petruckevitch A. A population
study of food intolerance. Lancet 1994;343:1127–30.
2. Woods RK, Abramson M, Bailey M, Walters EH. Inter-
national prevalences of reported food allergies and intol-
erances. Comparison arising from the European
Community Respiratory Health Survey (ECRHS)
1991–1994). Eur J Clin Nutr 2001;55:298–304.
3. Crowe SE, Perdue MH. Gastrointestinal food hypersensi-
tivity: basic mechanism of pathophysiology. Gastroenter-
ology 1992;103:1075–95.
4. Shanaham F. Food allergy: fact, fiction, and fatality. Gas-
troenterology 1993;104:1229–31.
5. Zar S, Benson MJ, Kumar D. Food-specific serum IgG4
and IgE titers to common food antigens in irritable bowel
syndrome. Am J Gastroenterol 2005;100:1550–7.
6. Astute. Statistics add-in for Microsoft Excel. DDU software
ver. 1.5. Leeds, UK: The University of Leeds, Old Medical
School, 1995.
7. Petrie A, Sabin C. Medical statistics at a glance, 2nd ed.
Oxford: Blackwell Science Ltd., 2000.
