Full text of DOI:10.1007/BF03344035

J. Endocrinol. Invest. 25: 442-446, 2002
The influence of parity on multinodular goiter prevalence in
areas with moderate iodine deficiency
M. Rotondi*, F. Sorvillo*, G. Mazziotti*, S. Balzano**, S. Iorio*, A. Savoia*, M. Piscopo*,
B. Biondi***, G. Amato*, and C. Carella*
*Institute of Endocrinology, Second University of Naples, **Department of Mathematics and Statistics,
***Institute of Endocrinology, University “Federico II” of Naples, Naples, Italy
ABSTRACT. Despite the observation that parity
may increase the risk of thyroid carcinoma, very
few studies have investigated the possible reper-
cussion of parity on thyroid benign pathology.
Recently, parity has been identified as one of the
factors contributing to a larger thyroid size in
healthy females. The aim of this work was to in-
vestigate a possible role for parity on the preva-
lence of multinodular goiter in iodine deficient ar-
eas. For this purpose, the reproductive histories
of 2 cohorts of women, normal (Group I, 235 cas-
es) and non-toxic multinodular goiter (NTMNG)
affected (Group II, 274 cases) were compared. All
subjects were euthyroid and had no previous his-
tory of thyroid function abnormalities. The num-
ber of full-term previous pregnancies (2.55 0.11
vs 1.77 0.10) and age (47.7 0.76 vs 42.3 0.83
yr) were found significantly higher (p<0.001) in
multinodular goiter (MNG) patients than controls.
Parity and age were found to be directly corre-
lated (p<0.001), nevertheless the partial correla-
tion coefficients demonstrated an independent
and statistically significant difference for both
variables between normal and NTMNG. Therefore,
the independent effects of parity and age were
further investigated. The effect of age on NTMNG
prevalence seems to be weaker, in fact significant
differences (p<0.001) for age between patients
and controls were detected only when the effect
of parity was absent (nulliparous), while with in-
creasing gestations the effect of age disappeared.
Our results indicate that age plays a minor role
compared to parity which can therefore be con-
sidered as a stronger risk factor. In conclusion, the
present study shows that, at least in iodine defi-
cient regions, non-toxic multinodular goiter wom-
en show a statistically significant higher parity rate
than healthy controls. Age may play a certain role
but only when additional stronger risk factors are
absent.
(J. Endocrinol. Invest. 25: 442-446, 2002)
©2002, Editrice Kurtis
INTRODUCTION
sufficient iodine supply, at least partly explains the
higher prevalence of thyroid disease in women than
in men. Several epidemiologic studies have demon-
strated that parity may increase the risk of thyroid
carcinoma (10) and a significant trend for increas-
ing risk with increasing parity was demonstrated in
a large cohort of women (11). However, very few
studies have investigated the role of parity in rela-
tion to benign thyroid disease, as the clinical trials
were limited to studies conducted in previously ir-
radiated individuals (12), autoimmune thyroiditis
(13) or thyroid nodules (14). In the latter report, a
higher prevalence of thyroid nodules with increas-
ing number of previous pregnancies in a region
with moderate iodine deficiency was reported (14).
More recently, we demonstrated that in compara-
ble conditions of iodine intake the goitrogenic ef-
fect of pregnancy is not fully reversible and parity
has been identified as one of the factors contribut-
When a pregnancy takes place in even moderate
iodine deficient regions, it is almost invariably ac-
companied by enlargement of the thyroid gland (1-
4). The increase in thyroid volume has been demon-
strated to be directly correlated to the severity of
iodine restriction (5, 6). Pregnancy-induced meta-
bolic changes may result in the development of goi-
ter in about 10% of pregnant women (7, 8) and it
has been reported that goiter can persist after pre-
gnancy in a significant number of cases (9). Clinical
evidence has supported the hypothesis that preg-
nancy, especially in regions with borderline or in-
Key-words: Parity, pregnancy, multinodular goiter, iodine deficiency.
Correspondence: Prof. Carlo Carella, Via Crispi 44, 80121 Napoli, Italy.
E-mail: Carlo.Carella@unina2.it
Accepted January 15, 2002.
442

Parity and multinodular goiter
Ultrasound examination
ing to a larger thyroid size in healthy females (15).
Difficulty in performing this kind of prospective
study together with the large series of patients
needed in order to draw relevant statistical conclu-
sions probably accounts for the scarce sources
available in this field. Although limited, these re-
ports have sustained the concept that at least in io-
dine deficiency conditions, parity may have certain
influences on the thyroid gland physiopathology,
suggesting the need for further investigations. In
this view, we decided to perform the present study
with the aim of investigating a possible role of par-
ity on the prevalence of multinodular goiter in io-
dine deficient areas.
A Toshiba 250 scanner was employed, using a 7.5 MHz trans-
ducer. All subjects were examined in the supine position, with
the neck in hyperextension. The total volume calculation was
performed by a previously described method (17), with in-
traobserver and interobserver variation of 4% and 6%, respec-
tively.
Statistical analysis
Statistical analysis was performed using SPSS software (SPSS, Inc.,
Evanston, U.S.A.). Sampling modality does not reflect goiter
prevalence in our region. Therefore, the 2 groups have been con-
sidered as samples of independent populations (normal/ goitrous),
for all the statistical analyses performed. Since the statistical aim
of this work was to compare the reproductive histories of normal
and goitrous females, results cannot be regarded as representa-
tive of a general phenomenon, but should be viewed as a com-
parison between the 2 populations. Relative frequencies of pre-
vious pregnancies for normal and goitrous women were calculat-
ed. Parity and age were compared between the 2 groups by
Student t test. The relationship between age and parity was as-
certained by Pearson’s correlation coefficient. To test the inde-
pendent effect of parity and age partial correlation coefficients
were computed. To further investigate the independent effect of
the 2 variables, t test for independent samples was used to com-
pare ages among groups with different parity status. A p value
<0.05 was considered statistically significant.
PATIENTS AND METHODS
In this study, data regarding the parity status for healthy females
and non-toxic multinodular goiter (NTMNG) affected patients
were mainly recollected from 2 recently published reports (15,
16). New cases were added either from healthy volunteers or
from people attending thyroid surgery for multinodular goiter.
The study group encompassed 509 females (235 normal and 274
multinodular goiter). Their mean SE ages were 42.3 0.83 and
47.7 0.76 yr for the 2 groups, respectively. All the women lived
in Naples, Italy, or the surrounding areas, a region known to pre-
sent moderate iodine deficiency, with usual urinary iodine levels
ranging 40-100 μg/d. In order to exclude any kind of thyroid func-
tion abnormalities, a serum FT3, FT4 and TSH evaluation was per-
formed in all cases. All 509 patients enrolled were euthyroid, thy-
roid autoantibodies were negative and had no previous history of
thyroid function disease. They had never performed iodine pro-
phylaxis. An accurate reproductive anamnesis was registered for
all patients either during regular follow-up consultations or when
performing the ultrasound evaluation. All cases with a pregnan-
cy in the previous 18 months or with previous history of miscar-
riages and abortions were taken into account to be excluded
from the study. Multinodular goiter was defined as the ultrasound
detection of an enlarged gland with more than one nodule of at
least 1.0 cm maximum diameter. The benign condition was as-
certained either by fine needle aspiration biopsy or histology.
Distinctive inclusion criteria for normal (group I) were: no detec-
table nodular lesion within a normal sized thyroid gland at ultra-
sound, no previous levothyroxine assumption, while for NTMNG
(group II) were: a thyroid volume calculated by ultrasound greater
than 26 ml, absence of thyroid malignancy. All subjects enrolled
met these criteria.
RESULTS
Age, parity status and thyroid parameters for nor-
mal and goitrous women are given in Table 1. The
influence of parity status on the relative frequen-
cies of the 2 populations (normal/goitrous) is clear-
ly shown in Figure 1. The distribution of relative fre-
quencies suggests that with increasing parity, the
percentages of normal and goitrous women de-
crease and increase, respectively. Both parity
(2.55 0.11 vs 1.77 0.10) and age (47.7 0.76 vs
42.3 0.83 yr) were found significantly higher
(p<0.001) in patients with multinodular goiter
(MNG) than controls. As expected, the 2 covariates
(parity and age) showed a significant correlation
(p<0.001). The results of partial correlations per-
Table 1 - Age, parity status and thyroid parameters (mean SE)
for normal and goitrous women.
Hormone assay
Variable
N
NTMNG
274
p value
Serum free T3, and free T4 (normal range 3.84-7.68 and 10.3-
23.2 pmol/l, respectively) were assayed by RIA with Lysophase
kits (Technogenetics kit set Milan, Italy). Intra- and interassay
variations and sensitivities were 2.9%, 4.7% and 0.8 pmol/l for
FT3 and 3.0%, 5.7% and 1.0 pmol/l for FT4, respectively. Serum
TSH levels (normal range within 0.3 and 3.5 mU/l) were investi-
gated by an ultrasensitive assay (DiaSorin kit set, Saluggia, Italy).
Intra- and interassay variability was 3.9 and 5.4% respectively;
sensitivity was 0.05 mU/l.
No. of cases
Age (yr)
235
42.3 ( 0.83)
1.77 ( 0.10)
47.7 ( 0.76)
2.55 ( 0.11)
<0.001
<0.001
No. pregnancies
Plasma TSH (mU/l)
FT4 (pmol/l)
FT3 (pmol/l)
1.4 ( 0.06)
16.3 ( 0.22)
4.5 ( 0.04)
1.3 ( 0.06)
15.9 ( 0.24)
4.7 ( 0.03)
NS
NS
NS
NS: non significant; NTMNG: non-toxic multinodular goiter.
443

M. Rotondi, F. Sorvillo, G. Mazziotti, et al.
35
N
significant older age (p<0.001) was found in pa-
tients, while with increasing parity such difference
was abolished.
NTMNG
28.5
28.6
30
25
20
15
10
5
27
25.9
20.1
DISCUSSION
17.5
In this study, we compared the reproductive histo-
ries of 274 NTMNG with those of 235 normal con-
trols to determine whether parity could affect the
subsequent development of nodular goiter in iodine
deficient areas. The results, obtained in iodine re-
stricted conditions, indicate that a higher parity char-
acterizes patients affected by multinodular goiter.
The topic of thyroid and pregnancy has received in-
creasing attention over the last decade with signifi-
cant advances in our understanding of the phys-
iopathology of the glandular adaptation to preg-
nancy (1). Most previously published papers investi-
gated the pregnancy repercussion on the thyroid
gland between early gestation and term or the im-
mediate post-partum period, while very few data are
available in which potential long-term repercussions
of pregnancy on the thyroid have been investigat-
ed. Even the most relevant epidemiologic studies,
aimed at examining the prevalence and incidence of
thyroid disease in the community as well as their ae-
tiological risk factors and natural history, have only
marginally investigated the possible role of parity
(18, 19). Exceptions are given by the demonstration
that parity could increase the risk of thyroid carcino-
mas with a significant trend for increasing risk with
increasing parity (11). Such trend was later ascribed
to biologic changes occurring during pregnancy as
no correlation was found between the number of
children and thyroid carcinoma in men (20). The lack
of any influence of parity on the prevalence of be-
nign thyroid adenomas was reported in 2 previous
studies both conducted in irradiated individuals (12,
21). Controversy in this field may easily be explained
by individual differences in the patients studied.
Conceivably inclusion of women with a proven
strong risk factor for thyroid disease (i.e. X-ray ex-
posure) resulted in decreases in estimates of the par-
ity related risks. The only clinical attempt to investi-
gate the influence of previous pregnancy on benign
thyroid pathology in an unselected population de-
monstrated that in iodine deficiency there is a high-
er prevalence of thyroid nodules with an increasing
number of previous pregnancies (14). Although sig-
nificant, such observation deserves a comment.
Indeed, it is known that the prevalence of thyroid
nodules increases with age, therefore a possible sta-
tistical bias might have been introduced by the fact
that even if the study group included only middle-
aged women (36-50 yr), we are not given any infor-
12.8
6.6
11.7
11.3
5.1
4.2
0
0
1
2
3
4
>5
Pregnancies (no.)
Fig. 1 - Relative frequencies for normal and goitrous women in
relation to the number of previous full-term pregnancies. With in-
creasing parity, the relative frequencies of normal and of wom-
en affected by non-toxic multinodular goiter (NTMNG) decrease
and increase respectively. N: normal; NTMNG: non-toxic multi-
nodular goiter.
formed after the adjustments for age and parity
demonstrated an independent and significant
(p<0.001 in both cases) difference for the 2 pa-
rameters between normal and NTMNG. Therefore,
in order to test the real importance of parity and
age considered independently, further analysis was
required. With this aim, a division of all cases on
the basis of their full-term pregnancies was per-
formed and potential differences for age, between
normal and goitrous, were evaluated. As shown in
Figure 2, only for nulliparous women a statistically
80
70
*
60
50
40
30
20
N
NTMNG
No. cases 67 47
0
29 18
1
66 73
2
45 69
3
13 32
4
9
31
ꢀ5
Pregnancies (no.)
Fig. 2 - Comparison of age between normal and goitrous wom-
en in relation to the number of their previous full-term preg-
nancies. Statistically significant differences (*p<0.001) could be
detected only for nulliparous women. N: normal; NTMNG: non-
toxic multinodular goiter.
444

Parity and multinodular goiter
REFERENCES
mation about possible differences in age between
the various groups, especially because the Authors
considered only the presence of thyroid nodules, ir-
respectively of their presence in the context of a nor-
mal or enlarged gland.
Since this type of study cannot be performed
prospectively, we decided to compare the number
of full-term pregnancies in women previously stud-
ied. Although the design of the present study was
retrospective, the results constitute the first clinical
confirmation that, at least in iodine deficient con-
ditions, the prevalence of NTMNG increases in re-
lation to the number of pregnancies.
1. Glinoer D. The regulation of thyroid function in pregnan-
cy: pathways of endocrine adaptation from physiology to
pathology. Endocr. Rev. 1997, 3: 404-433.
2. Berghout A., Wiersinga W. Thyroid size and thyroid func-
tion during pregnancy: an analysis. Eur. J. Endocrinol.
1998, 138: 536-542.
3. Glinoer D. What happens to the normal thyroid during
pregnancy? Thyroid 1999, 9: 631-635.
4. Glinoer D., Lemone M. Goiter and pregnancy: a new in-
sight into an old problem. Thyroid 1992, 2: 65-70.
5. Pedersen K.M., Laurberg P., Iversen E., et al. Amelioration
of some pregnancy-associated variations in thyroid func-
tion by iodine supplementation. J. Clin. Endocrinol. Metab.
1993, 77: 1078-1083.
The mechanisms by which parity would lead to
higher prevalence of disease can be reasonably ex-
plained. Iodine deficiency remains the main cause
for MNG. Several prospective clinical trials have
demonstrated that pregnancy even in a region with
marginal iodine deficiency may lead to a relative
iodine deficient state with potential repercussions
both on the mother (6) and the offspring (22).
Indeed, the pregnancy-induced metabolic adapta-
tions are compromised to different extents, relat-
ed to the severity of iodine restriction (5, 6). There-
fore, it seems plausible that women with more
pregnancies, viewed as a prolonged physiological
condition stimulating the thyroid gland (23), may
develop goiter more frequently, or at least facili-
tate its occurrence.
Despite the fact that no systematic iodine prophy-
laxis has ever been undertaken, goiter prevalence in
our region has decreased throughout the last 4
decades (24-26). Our results suggest that, besides
the improvement of nutritional factors, the significant
decrease in natality may also have played a role.
The effect of age on NTMNG prevalence has also
been considered. Our results suggest a weak effect
played by age, in fact significant differences for age
between patients and controls could be detected
only in nulliparous women. In other words, an old-
er age characterizes NTMNG women only when the
effect of parity is absent (nulliparous). With in-
creasing gestations the effect of age gradually dis-
appears as indicated by comparable ages between
the 2 groups and confirmed by the fact that for
more than 5 pregnancies, normal women are even
older than patients. This clearly indicates that age
plays a minor role compared to parity which can
therefore be considered a stronger risk factor.
In conclusion, the present study shows that in io-
dine deficient regions a statistically significant high-
er parity rate is found in NTMNG than in healthy
controls. Age can therefore be considered to weak-
ly affect goiter prevalence and only when additional
stronger risk factors are absent.
6. Glinoer D., De Nayer P., De Lange F., et al. A randomized
trial for the treatment of mild iodine deficiency during
pregnancy: maternal and neonatal effects. J. Clin. Endo-
crinol. Metab. 1995, 80: 258-269.
7. Caron P., Hoff M., Bazzi S., et al. Urinary iodine excretion
during normal pregnancy in healthy women living in the
Southwest of France: correlation with maternal thyroid pa-
rameters. Thyroid 1997, 7: 749-754.
8. Smyth P.P.A., Hetherton A.M.T., Smith D.F., Radcliff M.,
O’Herlihy C. Maternal iodine status and thyroid volume
during pregnancy: correlation with neonatal iodine intake.
J. Clin. Endocrinol. Metab. 1997, 82: 2840-2843.
9. Glinoer D., Lemone M., Bourdoux P., et al. Partial re-
versibility during late post-partum of thyroid abnormalities
associated with pregnancy. J. Clin. Endocrinol. Metab.
1992, 74: 453-457.
10. Preston-Martin S., Bernstein L., Pike M.C., Maldonado
A.A., Henderson B.E. Thyroid cancer among young wom-
en related to prior thyroid disease and pregnancies histo-
ry. Br. J. Cancer 1987, 55: 191-195.
11. Kravdal O., Glattre E., Haldorsen T. Positive correlation
between parity and incidence of thyroid cancer: new evi-
dence based on complete Norwegian birth cohorts. Int. J.
Cancer 1991, 49: 831-836.
12. Wong F.L., Ron E., Gierlowski T., Schneider A.B. Benign thy-
roid tumors: general risk factors and their effects on radia-
tion risk estimation. Am. J. Epidemiol. 1996, 144: 728-733.
13. Phillips D.I.W., Lazarus J.H., Butland B.K. The influence of
pregnancy and reproductive span on the occurrence of
autoimmune thyroiditis. Clin. Endocrinol. (Oxf.) 1990, 32:
301-306.
14. Struve C.W., Haupt S., Ohlen S. Influence of frequency of
previous pregnancies on the prevalence of thyroid nod-
ules in women without clinical evidence of thyroid disease.
Thyroid 1993, 3: 7-9.
15. Rotondi M., Amato G., Biondi B., et al. Parity as a size-de-
termining factor in areas with moderate iodine deficiency.
J. Clin. Endocrinol. Metab. 2000, 85: 4534-4537.
16. Rotondi M., Amato G., Del Buono A., et al. Post-interven-
tion serum TSH levels may be useful to differentiate pa-
445

M. Rotondi, F. Sorvillo, G. Mazziotti, et al.
tients who should undergo levothyroxine suppressive ther-
apy after thyroid surgery for multinodular goiter in a re-
gion with moderate iodine deficiency. Thyroid 2000, 10:
1081-1085.
X-irradiated in infancy for enlarged thymus glands. Radiat.
Res. 1993, 134: 217-223.
22. Aghini Lombardi F., Pinchera A., Antonangeli L., et al.
Mild iodine during fetal/neonatal life and neuropsycho-
logical impairment in Tuscany. J. Endocrinol. Invest.
1995, 18: 57-62.
17. Rasmussen S.M., Hjorth L. Determination of thyroid vol-
ume by ultrasonic scanning. J. Clin. Ultrasound 1974, 2:
143-147.
23. Glinoer D. Pregnancy and iodine. Thyroid 2001, 11: 471-
18. Vanderpump M.P.J., Tunbridge W.M.G., French J.M., et
al. The incidence of thyroid disorders in the community: a
twenty-year follow-up of the Whickham Survey. Clin.
Endocrinol. (Oxf.) 1995, 43: 55-68.
481.
24. Report of the Subcommittee for the study of endemic
goitre and iodine deficiency of the European Thyroid
Association. Goitre and iodine deficiency in Europe. Lancet
1985, 1: 1289-1293.
19. Wang C., Crapo L.M. The epidemiology of thyroid disease
and implications for screening. Endocrinol. Metab. North
Am. 1997, 1: 189-218.
25. Gutekunst R., Sciba P.C. Goiter and iodine deficiency
in Europe. The European Thyroid Association report as
update in 1988. J. Endocrinol. Invest. 1989, 12: 209-
220.
20. Glattre E., Kravdal .. Male and female parity and risk of
thyroid cancer. Int. J. Cancer 1994, 58: 616-617.
21. Shore R.E., Hildreth N., Dvoretsky P., Pasternack B.,
Andresen E.P. Benign thyroid adenomas among persons
26. Delange F. The disorders induced by iodine deficiency.
Thyroid 1994, 4: 107-128.
446