Evidence of cyclic alterations of thyroid size during the menstrual cycle in healthy women

Laszlo Hegedus, M.D., Steen Karstrup, M.D., and Nini Rasmussen, M.D.

Her/ev, Denmark

Serum levels of thyroxine, triiodothyronine, triiodothyronine resin uptake, thyroxine-binding globulin, thyrotropin, and ultrasonically determined thyroid gland volume were investigated weekly in 11 healthy women during a normal menstrual cycle, and in five healthy men once a week for 5 consecutive weeks. In the men the thyroid volume was unaltered. In the women, however, a mean variation in thyroid volume of approximately 50% between minimum values (15.4 ± 3.1 ml, day 9) and maximum values (24.4 ± 4.8 ml, day 23) was found (p < 0.01), although no significant differences in the other thyroid variables could be demonstrated. The menstrual cycle seems to be associated with cyclic alterations of thyroid size in healthy women unexplained by alterations in thyroid function variables. This information should be taken into account when goiter frequency, goitrogenic action of drugs, and goiter treatment effects are evaluated. (AM J OBSTET GYNECOL 1986;155:142-5.)

Key words: Thyroid hormones, thyrotropin, thyroid volume, goiter, menstrual cycle, ultrasound

It is a common clinical experience that many women have noticed variations in neck circumference during the menstrual cycle. No data are available on the influ­ence of the menstrual cycle on thyroid gland size, prob­ably because of the lack of sufficiently sensitive objective methods for this evaluation. The aim of the present study was to determine whether thyroid gland volume, ultrasonically determined,' showed any changes during the menstrual cycle in healthy women.

Subjects and methods

Subjects. Included in the study were the following subjects: (I) Eleven women with a median age of 26 years (range, 18 to 42 years) and a median body weight-on day 2 of their menstrual cycle-of 54 kg (range, 52 to 64 kg). All had normal menstrual cycles, which were regular for at least 6 months before the study. The investigations were begun on the second day of menstrual bleeding and terminated on the second day of the following menstrual bleeding. Thyroid vol­ume determination and blood sampling were per­formed once a week for a whole menstrual cycle (days 2, 9, 16, and 23 and day 2 in the following cycle). (2) Two of these women had their thyroid volume deter­mined almost every day during a complete menstrual

From the Department of Internal Medicine and Endocrinology F and the Ultrasonic Department, HerlevHospital, University of Copen­hagen.

Received for publication April 29, 1985; revised January 23, 1986; accepted january 27, 1986.

Reprint requests: Laszlo HegedUs, M.D., Department of Internal Medicine and Endocrinology F, Herlev Hospital, DK-2730 Her­lev, Denmark.

142

cycle. (3) Five healthy young men with a median age of 32 years (range, 28 to 34 years) and a median body weight of 66 kg (range, 62 to 83 kg) were also studied. Starting on randomly chosen days and continuing once a week for 5 weeks, thyroid gland volume was deter­mined. No blood samples were drawn in these subjects.

Upon entry into the study none of the subjects had known present or past thyroid disease or a detectable goiter. None of the women had been pregnant within the last year, and none of the subjects were taking any kind of medication including oral contraceptives.

Methods. The presence of a goiter was defined as a visible and/or palpable thyroid gland. The ultrasonic scanning procedure and the total volume calculation were performed as previously described with a com­pound scanner (type 1846, Briiel and Kjaer, Naerum, Denmark).' The transducer was a 5.5 MHz type me­chanically focused at 1 to 3 cm. The scanning procedure and volume calculation as well as the accuracy and pre­cision of the method were reported previously.' Thy­roid volume measurements were always performed on the day of thyroid function investigations at 4 PM. All volume determinations in one subject were performed by the same investigator without knowledge of the sub­ject's menstrual cycle phase. The precision of the ul­trasonic method was 5.1 % in the present study, ex­pressed as the coefficient of variation on double deter­minations. Measurements of thyroid volume in four normal men three times a day on 3 consecutive days showed a mean coefficient of variation of 5.7% in day­to-day measurements and of 4.8% in measurements within the same day.

Serum thyroxine, triiodothyronine, triiodothyronine

Volume 155 Number 1

Q)

E ~

g "t:I

i :2 E! >-= ~S

30

26

22

18

14

10

L< ,

Thyroid volume and menstrual cycle 143

• .... " ....

" .... .... ....

"1 • p< 0.01

,---., 2 6 10 14 18 22 26 28 2

Menstrual cycle (days)

Fig. 1. Thyroid gland volume (mean ± SD) in 11 healthy women during a normal menstrual cycle. Investigations were performed once a week, and mean thyroid volume was significantly higher on days 16 and 23 than on days 2 and 9 (p < 0.01).

36

32

Q) 28 E ~

g 24 "t:I

te= CIS :g 20 E! >-

.J::. I-

16

12 l' "~----~i----~i----~----~--~i----~----~i-Fj'~

5 9 13 17 21 25 29 1

Menstrual cycle (days)

Fig. 2. Thyroid gland volume determined almost every day during a normal menstrual cycle in two healthy women.

resin uptake, and thyrotropin were determined as pre­viously described. 1.2 The interassay variation for serum thyrotropin determination was 7.2%, and the intra-as­say variation was 5.8% in the range of serum thyrotro­pin measured. The detection limit was 0.3 fLU/ml. Thy­roxine-binding globulin was measured by rocket im­munoelectrophoresis.3 The free thyroxine index and free triiodothyronine index were calculated as total hormone concentration times triiodothyronine resin uptake and given in arbitrary units. Serum was stored at - 20° C, and analyzed in random order in the same assay series.

Results were examined by one-way analysis (Fried­man's test for related samples). If this test indicated that some differences existed, multiple comparisons were performed according to Conover.4 Probability val­ues of <0.05 were regarded as significant.

Results

Among the women the median body weight was 54 kg (range, 52 to 64 kg) on day 2 and 56 kg (range, 54 to 66 kg) on day 23 (p < 0.(5). A significant increase in thyroid gland volume during a normal menstrual cycle was observed (Fig. 1). Thus mean thyroid volume

144 HegedOs, Karstrup, and Rasmussen

26

July, 1986 Am J Obstet Gynecol

---- . ------.. ---- •

~ •

10

1 2 3

Time (weeks)

4 5

Fig. 3. Thyroid gland volume determined once a week for 5 consecutive weeks in five healthy men.

Table I. Thyroid function variables during the normal menstrual cycle in 11 healthy women (mean ± SD)

Cycle day

2 9

16 23 2 (of fol­low-

ing cycle)

Thyroxine (nmollL)

95.5 ± 13.8 102.7 ± 10.8 94.6 ± 7.9 95.6 ± 11.3 96.2 ± 9.4

Triiodothyronine (nmollL)

2.1 ± 0.4 2.1 ± 0.3 2.1 ± 0.5 2.1 ± 0.3 2.2 ± 0.4

Triiodothyronine resin uptake

(arbitrary units)

1.00 ± 0.09 1.01 ± 0.02 1.00 ± 0.02 0.99 ± 0.02 0.99 ± 0.04

varied approximately 50% between minimum (15.4 ±

3.1 ml, day 9) and maximum values (24.4 ± 4.8 ml, day 23) (p < 0.01, Fig. I). Values were significantly higher on day 16 and 23 than on day 2 and 9 (p < 0.01, Fig. I). No significant difference was found between thyroid volume at the start (day 2) or at the termination of the investigation (day 2 of next cycle). None of the women had a clinically detectable goiter at any time during their menstrual cycle. In the two women inves­tigated every day during a whole cycle, maximum thy­roid volume was observed on days 22 to 23 and 25 to 27, respectively (Fig. 2). No such pattern could be ob­served in the men (Fig. 3). The mean alterations in thyroid volume in the men were of the same magnitude (4.6%) as the previously reported day-to-day variation of this parameter (5.7 % ).

No significant alterations in any of the other thyroid variables could be demonstrated in the women during the menstrual cycle (Table I).

Comment The present study verifies the experience of a men­

strual cycle-related cyclic alteration of thyroid size. The reason for this information not hitherto being available

Thyroxine-binding globulin

(arbitrary units)

104.7 ± 14.2 107.9 ± 14.0 107.2 ± 12.1 105.9 ± 14.4 105.8 ± 13.7

Free thyroxine index

(arbitrary units)

96.9 ± 15.5 103.5 ± 13.3 94.3 ± 11.3 94.7 ± 14.3 96.4 ± 15.0

Free triiodothyronine index

(arbitrary units)

2.12 ± 0.56 2.15 ± 0.30 2.12 ± 0.49 2.05 ± 0.23 2.17 ± 0.49

Thyrotropin (p-Ulml)

1.9 ± 0.8 2.3 ± 0.7 2.1 ± 0.8 2.4 ± 1.0 2.0 ± 0.8

is best explained by the lack of sufficiently sensitive methods for this evaluation. Furthermore, none of the women had a clinically detectable goiter at any time during the investigation, and thyroid volume alter­ations mainly occurred within the normal range.! Thy­roid volume is closely related to body weight,! so an increase in body weight in the second half of the men­strual cycle could in part be the explanation for this increase in thyroid volume. The present study does not allow speculations on the nature of this increase. How­ever, the abrupt return within a few days to normal size around the onset of the menstrual bleeding makes it most plausible to speculate regarding altered vascu­larity of the thyroid gland rather than hypertrophy.

It has been reported that female rats have cyclic vari­ations in thyrotropin and thyroid activity correlating with the estrous cycle.5 Some have found an increase in thyroid weight (hyperplasia) after administration of estrogen.6 Prolonged treatment, however, resulted in involution of the thyroid gland. Others have not found such an increase in thyroid weight. 7 Previous studies, also in animals, have shown that the action of estrogen on the thyroid was not mediated via the pituitary, and it was assumed that estrogen acted directly on the thy-

Volume 155 Number 1

roid.B.

9 The present study and other studieslo. 1\ sup­

port this concept, since no alterations in serum thyro­tropin levels could be demonstrated. In addition Carl­son et al. 12 found unaltered basal and thyrotropin-re­leasing hormone-stimulated serum thyrotropin values in healthy males treated with estrogen. It can; however, not be ruled out that the use of a new ultrasensitive thyrotropin method and/or more frequent blood sam­pling could have disclosed significant alterations in se­rum thyrotropin level.

It is well established that estrogens-synthetic as well as genuine-stimulate the production of thyroxine­binding globulin. \3 This normally leads to an increase in serum thyroxine and triiodothyronine levels but free thyroxine and free triiodothyronine levels are usually unchanged. This mechanism of estrogen-stimulated thyroxine-binding globulin production has been sug­gested as partly responsible for the goitrogenic effect of pregnancy. I< The present study was, however, not able to demonstrate alterations in serum thyroxine­binding globulin concentration or the other thyroid variables except thyroid volume. Furthermore, recent investigations have questioned this,l5· 16 and it is impor­tant to stress that no objective studies addressing the effect of pregnancy on thyroid size are available. In acute reversible liver disease (hepatitis) there is a re­versible increase in serum thyroxine-binding globulin (due to destruction ofhepatocytes) and thyroid volume, but serum thyrotropin level is unchanged. 17 The mech­anism underlying thyroid volume alterations during and after acute liver disease was not clarified, possibly again the result of infrequent blood sampling and/or lack of a supersensitive thyrotropin method.

The menstrual cycle is associated with changes in a number of hormonal systems such as aldosterone, 18 prolactin,19 catecholamines,2° progesterone," I and va­sopressin,22 but the interrelationship between these hormonal systems and thyroid function and size is not clarified.

In conclusion, a cyclic alteration in thyroid size unex­plained by alterations in thyroid hormones or thyro­tropin has been demonstrated. If these changes in thy­roid gland volume during a normal menstrual cycle are not considered, they might constitute a pitfall in the evaluation and comparison of the goiter frequency in different populations. Furthermore, if these alterations are also present in subjects with goiter, the evaluation of e.g. treatment results after thyroxine23 or tri­iodothyronine24 treatment of diffuse nontoxic goiter may be heavily biased if not corrected for this finding.

REFERENCES 1. Hegedus L, Perrild H, Poulsen LR, et al. The determi­

nation of thyroid volume by ultrasound and its relation­ship to body weight, age, and sex in normal subjects. ] Clin Endocrinol Metab 1983;56:260-3.

2. Hegedus L. Decreased thyroid gland volume in alcoholic

Thyroid volume and menstrual cycle 145

cirrhosis of the liver. ] Clin Endocrinol Metab 1984;58:930-3.

3. Nielsen HG, Buus 0, Weeke B. A rapid determination of thyroxine-binding globulin in human serum by means of the Laurell rocket immunoelectrophoresis. Clin Chim Acta 1972;36:133-8.

4. Conover W]. Practical non parametric statistics. 2nd ed. New York:] Wiley & Sons, 1980:299-305.

5. Boccabella AV, Alger EA. Quantitative variations in serum thy~otropin levels during the estrous cycle of the rat. En­docrinology 1967;81: 121-4.

6. Desclin L, Ermans AM. Action de I'hormone folliculaire sur I';lctivite thyreotrope de I'hypophyse chez Ie rat. Compt Rend Soc Bioi 1950;144:1277-9.

7. Feldman ]D. Effects of estrous and estrogen on thyroid uptake of 1311 in rats. Endocrinology 1956;58:327-37.

8. Bithell ]F, Brown-Grant K. An experimental analysis of the effects of estrogen on the thyroid gland of castrated adult male rats.] Endocrinol 1968;40:397-408.

9. Feldman ]D. Effects of estrogen on thyroidal iodide trap­ping and conversion of inorganic 131 1 to protein-bound "'I. Endocrinology 1956;58:289-92.

10. Weeke], Hansen AaP. Serum TSH and serum T3 levels during normal menstrual cycles and during cycles on oral contraceptives. Acta Endocrinol (Copenh) 1975;79:431-8.

II. Lemarchand-Beraud T, Vannotti A. Relationships be­tween blood thyrotrophin level, protein bound iodine and free thyroxine concentration in man under normal phys­iological conditions. Acta Endocrinol (Copenh) 1969; 60:315-26.

12. Carlson HE, Jacobs LS, Daughaday WHo Growth hor­mone, thyrotropin, and prolactin responses to thyrotro­pin-releasing hormone following diethylstilbestrol pre­treatment.] Clin Endocrinol Metab 1973;37:488-90.

13. Briggs MH. Biochemical basis for the selection of oral contraceptives.] Gynaecol Obstet 1979;16:509-17.

14. Drury MI, Sugrue DD, Drury RM. A review of thyroid disease in pregnancy. Clin Exp Obstet Gynecol 1984; II :79-89.

15. Levy RP, Newman DM, Rejall LS, Barford DAG. The myth of goiter in pregnancy. AM ] OBSTET GYNECOL 1980; 137:701-3.

16. Long T], Felice ME, Hollingsworth DR. Goiter in preg­nant teenagers. AM] OBSTET GYNECOL 1985; 152:670-4.

17. Hegedus L. Thyroid gland volume and thyroid function during and after acute hepatitis infection. Metabolism (in press).

18. Sundsfjord ]A, Aakvag A. Variations in plasma aldoste­rone arid plasma renin activity throughout the menstrual cycle with special reference to the pre-ovulatory period. Acta Endocrinol (Copen h) 1973;75:499-508.

19. Horrobin DF, Lloyd I], Lipton A, Burstyn PG, Durkin N, Muiruri KL. Actions of prolactin on human renal func­tion. Lancet 1971;2:352-4.

20. Kuchel 0, Hamet P, Cuche ]L, Tolis G, Fraysse], Genest J. Urinary and plasma cyclic adenosine 3',5'-monophos­phate in patients with idiopathic edema.] Clin Endocrinol Metab 1975;41:282-9.

21. O'Brien PMS, Selby C, Symonds EM. Progesterone, fluid and electrolytes in premenstrual syndrome. Br Med ] 1980;280: 1161-3.

22. Spruce BA, Baylis PH, Burd], Watson MJ. Variation in osmoregulation of arginine vasopressin during the hu­man menstrual cycle. Clin Endocrinol (Oxf) 1985;22: 37-42.

23. Hansen]M, Kampmann], Madsen SN, et al. L-thyroxine treatment of diffuse non-toxic goitre evaluated by ultra­sonic determination of thyroid volume. Clin Endocrinol (Oxf) 1979;1:1-6.

24. Perrild H, Hansen ]M, Hegedus L, et al. Triiodothyron­ine and thyroxine treatment of diffuse non-toxic goitre evaluated by ultrasonic scanning. Acta Endocrinol 1982; 100:382-7.