Menstrual blood loss and body iron stores in Brazilian women

CONTRACEPTION

MBNBTRUAL

ATL Andrade,

BLOOD LOB8 AND BODY IRON STORES IN BRAZILIAN WOMEN

JP Souza, ST Shaw, Jr., EM Belsey and PJ Rowe

Universidade Federal de Juiz de Fora, Brazil The University of Chicago, USA NH0 Special Programme of Research, Development and Research Training in Human Reproduction, Switzerland

Abstraat Menstrual blood loss (MBL) studies are relevant for developing world women as this could be an important cause of anemia. Whenever a contraceptive method is to be used by such women, consideration should be given to the method which least affects the volume of MBL. In 309 women considered as clinically healthy, MBL, serum ferritin, serum iron and hemoglobin levels were measured: a mean MBL of 23 ml was found. height and previous oral Age, weight, contraceptive use did not affect MBL. Higher parity women may have higher MBL levels but their hematologic indices are not altered. While body iron stores (as judged by serum ferritin levels) are depleted in women who bleed more than 60 ml per cycle, clinical anemia may not be present until their blood loss exceeds 80 ml per menstruation. Brazilian women who lose more than 60 ml of menstrual blood associated with multiple pregnancies without adequate iron supplementation may have a depletion of their body iron stores.

Reprint requests de Fora, Brazil

to Dr. ATL Andrade,

Caixa Postal

328, Juiz

Submitted for publication November 6, 1990 Accepted for publication January 9, 1991

MARCH 1991 VOL. 43 NO. 3

241

CONTRACEPTION

Introduotion Studies of menstrual blood loss (MBL) in the developing world are highly relevant considering the high incidence of low body iron reserves due to poor nutritional status, infestations and multiple pregnancies parasitic unaccompanied by dietary iron supplementation, with serious consequences for the women and their offspring. Long-term use of contraceptive methods such as hormonal methods and intrauterine devices may also affect MBL and consequently alter body iron stores. Women's subjective evaluation of blood loss is not reliable and very frequently the gynecologist must have objective measurements to elucidate pathological processes. Data on MBL and hematologic indices have been available for Sweden (1,2), the United Kingdom (3), Mexico (4-6), USA (7) and China (8). The lack of data on MBL and related body iron status in healthy Brazilian women prompted this study. Subjects and Methods Three-hundred-and-nine women, aged 15 to 48 years, considered as clinically healthy and suitable to have an intrauterine device or a vaginal ring inserted for contraception or to serve as non-contracepting control subjects, after complete physical, hematologic and cytologic examinations, had their MBL, hemoglobin (hgb), serum iron (SI) and serum ferritin (SF) determined. These assays were performed in women who had not used oral contraceptives in the preceding 45 days, or were at least 45 days after pregnancy termination. Women were instructed to carefully collect their menstrual tampons (Ob tampons) and bring them to the laboratory in opaque plastic bags as soon as bleeding ended. As the women brought their tampons, a sample of peripheral blood was drawn from their arm at fasting, in the morning. The study consisted of recording their age, weight, height, parity as well as history of previous hormonal contraceptive use and date of the end of their last pregnancy. The only previous hormonal method used was oral contraception. Menstrual blood loss was quantified according to the technique first described by Hallberg and Nilsson (8), adapted by Shaw (10) and modified by Newton et al. (11). Hemoglobin was assayed by the cyanmethemoglobin technique, serum iron was determined by the method of the International Committee for the Standardization of Hematology, and serum ferritin was measured by radioimmunoassay using Fer-Iron serum ferritin kits, Ramco Laboratories, Houston, Texas, USA.

242

MARCH 1991 VOL. 43 NO. 3

CONTRACEPTION Statistical

Methods

The distribution of MBL, serum iron and ferritin was found to be skewed and their values were therefore logThe ponderal index (Quetelet transformed before analysis. index) was obtained by the formula: weight (kg)/height (m) squared. All comparisons were made by means of the Chi-square test or t-test. The coefficient of correlation was used to determine the strength of the relationship between pairs of variables. Results Brazilian women in this study had the following indices: 29.4 years of age (range: 15-48), parity of 2.5 (range: O-13), MBL of 23 ml (range: l-l-155.8), blood hemoglobin of 13.2 g/d1 (range: 7.9-15.5), serum iron of 72.2 @g/d1 (range: 7.0-166.6) and serum ferritin of 23 ng/ml (range: 2.5-140.0). Age does not seem to affect MBL, hemoglobin, serum iron or serum ferritin indices. The first impression is that Brazilian women have hematologic values. However, apparently normal among subjects who are judged clinically healthy, it was found that 72 subjects (25.7%) had low levels of serum ferritin (<15 ng/ml), 33 subjects (16.7%) had low serum iron (~50 ug/dl) and 24 subjects (7.8%) had low hemoglobin levels (~12 g/dl). When parity is considered, this factor also does not seem to alter MBL. Rven though higher parity women show a tendency to have lower serum ferritin levels (Table I), this is not statistically significant. When MBL is grouped by intervals (Table II), it can be seen that hemoglobin, serum iron and serum ferritin levels decrease as the women bleed more. Results show that serum ferritin falls to very low levels (~15 ng/ml) when subjects bleed more than 61 ml, while with serum iron, only those who lost more than 81 ml of blood had values below 50 pg/dl. With regard to hemoglobin levels, even though the decrease is significant among MBL intervals, few subjects had anemic levels. The profiles of the heavy bleeders (greater than 60 ml of MBL) were checked against the normal bleeders with respect to age, weight, height and parity (Table III). The only significant difference was that heavy bleeders had a higher parity (3.6) when compared to normal bleeders (2.4).

MARCH 1991 VOL. 43 NO. 3

243

Table I. Parity, menstrual blood loss and serum ferritin in Brazilian women ____________1___1_____________________I_~~~~~~~~~~~~~~~~~~ Menstrual Para

N

Blood loss*

Serum

95% Confidence

Mean

N

ferritin* 95x Confidence

Mean

(Ml) Interval (ns/ml) Interval ___~_____-_______~-___I_________________~~~~~~~~~--~~---~~ 0

25

20.9

(15.5,28.8)

18

35.5

(22.9,55.1)

i

82

21.5

(18.3,25.3)

75

23.9

(20.4,28.1)

2

93

21.8

(18.3,26.0)

85

21.7

(18.2,25.9)

3

43

23.5

(18.2.30.2)

41

23.2

Ci8.5,29.2)

=)4

66

27.3

(22.6,33.0)

61

20.9

(i7.4,25.2>

23.0

<21.0,25.1)

23.0

(21.1,25.2)

TOTAL

*

Log

309

before analysis - Correlation 0.104, serum ferritin= 0.096

transformed

coefficient:

280

ML=

Table II. Nemtrual blood loss intervals, blood hemoglobin, eerw iroP and sew feKititi ----_-__--l_--_l---_P P--s---Interval (111

m__--

(g/dl)

--_-

Serue Ferritin

Sew iron

Blood Hemslobin !I) Nean 95111 Confidence

(N) Nean 95%Confidence

(N) Wean 9% Confidence (ug/dl)

Interval

__----_----__c

(ng/dl)

Interval

_--~-__-------_-

Interval

=(20

130 13.3

(13.2, 13.5)

82

788

(72.6, 85.4)

21 - 40

95 13.5

(13.3, 13.7)

58

75.6

(68.2,

83.8)

82 23.4

(20.2,

27.2)

I

SO 12.8

(12.6,

13.1)

32

57.3

(4b.S,

70.6)

46 10.6

(14.6,

23.7)

24

(12.6,

13.51

i8

n.9

(65.6, 87.9)

22 14.5

(10.3, 20.3)

47.3

cn.t,

82.5)

P 10.6

( 5.7, 19.7)

198 72.2

(67.9,

76.0)

260 23.0

(21.1, 25.2)

- 40

bi - 80

+Ml

rnnu

13.0

io 12.0

(10.9, 13.11

30913.2 (13.1, 13.3)

8

121 28.5

(25.1, 32.3)

_____________~______1__1____1__1__m--_--m e LOStramfoiled before atUlYSiS; correlation Coefficient = +.29S

244

MARCH 1991 VOL. 43 NO. 3

CONTRACEPTION Table

XXI. Profiles

of normal and heavy bleeders

-___________________~~~~~~~~-~~~~~~~~~~~~~~~~-~~~~~~~~~~-Normal (=(60

bleeders ml)

Heavy

bleeders

061

N = N = 275 ____________________~~~~~~~~~~~~~~~~~~~~~~~~~_~~~~~__~~~__ llean 29.2

6.5

55.9

9.7

Age

Weight( Height

kg) (cm)

Quetelet

+, SD

Uean 30.6

42.0

t

Sign

ml)

OG

34

diff

SD 6.4

8.8

NS

NS

159.3

6.0

157.0

6.R

NS

23.1

3.4

17.0

3.1

NS

Index

1.9 3.6 2.8 P(O.05 Parity 2.4 ____________________~~~~~~~~~~~~~~~~~~~~~~~~~_~~~~~~~~~~~_

Comparing normal and heavy bleeders in relation to time since last pregnancy, in intervals of 12 months (up to 54 statistically significant no were there months), Concerning previous oral contraceptive use, differences. comparing never users with users within the last 6 months and with users of more than 6 months previously, there were no statistically significant differences between normal and heavy bleeders. indices with the oral Comparing hematologic significant statistically categories, no contraceptive differences were found, even though there appears to be a tendency for lower MBL levels in women who had previous oral There were no statistically contraceptive use (Table IV). significant differences in hemoglobin, serum iron or serum ferritin levels among all oral contraceptive categories. Discussion blood loss reduces hematologic Excessive menstrual This fact gains indices and is a cause of clinical anemia. for third world women with poor considerable importance nutrition.

MARCH 1991 VOL. 43 NO. 3

246

Table

IV.

contraceptive

Previous

MBL

OC

US@

indices

and hematologic (OC) use

< 6 months

>6

by history

months

of

oral

Never-used

N

Mean

N

Mean

N

mean

1.31

22.5

I.21

21.5

56

27.6

13f

13.3

121

13.2

56

13.2

80

74.1

75

72.0

43

69.3

22.9

44

23.8

1___..1__-__~_1_____~~~~~~~~~~~~~~~~.~~~~~~~~~.~~~~~~~~~.~~.-~-~~

MBL(m1) Hgb(

g/dl)

SI(ug/dl)* SF(ng/ml.)+

*

Log

transformed

125

ii0

22.9

before

analysis

Menstrual Blood Loss The range of MBL is different for many parts of the world: in Sweden it varied from 1.6 to 199.7 ml (1,9). In Mexico from 1.3 to 143.5 ml (6), and these results are comparable with those of the present study (1.1 to 155.8 ml). In China, however, the measurements were much higher, from 4.1 to 273.6 ml (8). The mean values of MBL were also different: in American women mean MBL was 30 ml (7), in Sweden 38.5 ml (l), in Mexico 35.1 ml (6), in the UK 37.5 ml (3) and in China 54.2 ml (8). In the present study, the mean MBL was 23.022.2 ml, which was lower than all other reports, and less than half that of Chinese women. The reason for such discrepancies is not known at the present time but genetic, nutritional and racial differences are possibly the main causes (2).

Age did not affect MBL or the hematologic indices studied, and this confirms the Mexican reports (4-6).

246

MARCH 1991 VOL. 43 NO. 3

CONTRACEPTION

Parity did not alter MBL in Mexican women (5), but the number of subjects grouped for each parity category was small in that study. Similarly, parity did not seem to affect MBL in Chinese women (S), but the subjects had very low parity (52% were nulliparous and 42% were primiparous). In Brazilian women there was a tendency for increasing MBL from the nulliparous to the grand multiparae (Table I), On the other but this was not statistically significant. hand, when comparing the parity of women who bled less than 60 ml per menstruation (2.4) with those who had MBL greater statistically than 60 ml the difference wa8 (3.6), significant (pCO.05). Comparing Tables I and III, it can be suggested that higher parity women may have higher MBL levels. For serum ferritin, there is a tendency for decreased values as the subjects have higher parity, but this, also, was not statistically significant. Weiaht

and Heiaht

In contrast with the Chinese study (S), height did not influence MBL in Brazilian women.

weight

and

&J&z olocric In Mexican women (4), in Chinese subjects (8) and in Swedish women (l), values of hemoglobin decreased when MBL was greater than 81 ml, and this was confirmed in the present study. Serum iron measurements were also below normal levels (50 kg/dl) when subjects bled more than 81 ml per menstrual cycle. Serum ferritin levels were inversely related to the amount of blood loss, and this confirmed other studies In Brazilian women, serum ferritin ranged from 2.5 (7,s). to 140 ng/ml and 25.7% of them had below normal levels (15 This level was reached when subjects had MBL ng/ml) greater than 60 ml, as similarly occurred in Chinese (8) and American women (7). As it is known that serum ferritin is a reliable index of body iron stores and its depletion occurs before the patient presents clinical anemia as demonstrated by low hemoglobin or serum iron, this study suggests that MBL above 60 ml is pathological for Brazilian women or, at least, potentially so.

MARCH 1991 VOL. 43 NO. 3

247

CONTRACEPTION

Previous Preanancv and MBL Time since the end of the last pregnancy did not influence the amount of MBL, and this confirmed a previous report (12). Previous Oral Contraceotive Use It is widely known that the present use of oral contraception reduces the amount of menstrual blood flow. In a previous report (12) comparing 52 subject who had prior 37 contraceptive use (more than days before oral measurements) with a mean MBL of 21.9 ml, with 36 neverusers who had a mean MBL of 47.1 ml, it was found that this difference was statistically significant. In the present study, however, 252 prior oral contraceptive users had a mean of 23 ml, while 56 never-users had a MBL of 27.6 ml, and this difference was not statistically significant. Although there may be a suggestion that previous oral contraception could decrease MBL, this could not be confirmed in this larger study. In conclusion, Brazilian women do seem to have lower MBL than other nationalities, and this may be due to genetic, ethnic and/or nutritional differences. Age, weight, height, time since last pregnancy and previous oral contraceptive use do not affect MBL. It is suggested that higher parity women have higher MBL levels, while parity per se does not alter the hematologic indices studied. While body iron stores are depleted in women who bleed more than 60 ml per cycle (by serum ferritin levels), clinical anemia may not be present until they bleed more than 80 ml per menstruation. For Brazilian women of high parity, the loss of more than 60 ml of menstrual blood unaccompanied by adequate iron supplementation during their pregnancies may be a cause of low body iron stores. Acknowledgement This investigation received financial support from the NH0 Special Programme of Research, Development and Research Training in Human Reproduction. References 1.

Hallberg L, Hogdahl AM, Nilsson L, Rybo G. Menstrual blood loss - a population study variation at different ages and attempts to define normality. Acta Obstet Gynaecol Stand 1966;45:320-51.

2.

Rybo G. Menstrual blood loss in relation to parity and menstrual pattern. Acta Obstet Gynaecol Stand 1966;45(suppl 7):25.

MARCH 1991 VOL. 43 NO. 3

CONTRACEPTION 3.

Billewica WT, Back AE. Thomson AM, Cole SK, Haematological characteristics and menstrual blood loss. J Obstet Gynaecol Brit Commonw 1972;79:994.

4.

Relation Aznar R, Pedron N, Gallegos A, Gonzalez, M. de1 sangrado menstrual y 10s valores de hemoglobina venosa. Ginec Obstet Mex 1979;46:99-105.

5.

Aznar R, Pedron N, Gallegos A, Gonzalez M. Relation de la paridad con el sangrado menstrual. Ginec Obstet Mex 1980;47:95-100.

6.

Pedron N, Gallegos A, Gonzalez M, Aznar R. Valoracion de la cantidad de1 sangrado menstrual en la mujer mexicana. Ginec Obstet Mex 1979;45:429-37.

7.

Shaw ST Jr, Aaronson DE, Moyer DL. Quantitation of menstrual blood loss -- Further evaluation of the alkaline hematin method. Contraception 1972;5:497-513.

8.

Gao Ji, Zeng Su, Sun Bo-ling, Fan Hui-min, Han Li-hui. Menstrual blood loss and haematologic indices in healthy Chinese women. J Reprod Med 1987;22:822-29.

9.

Hallberg L, Nilsson L. Determination of menstrual blood loss. Scandinav J Clin Lab Invest 1964;16:24448.

10.

Shaw ST Jr. On quantifying menstrual blood Contraception 1977;16:283-85.

11.

Newton J, Barnard G, Collins W. A rapid method for menstrual blood loss using measuring automatic extraction. Contraception 1977;16:269-82.

12.

Andrade ATL, Souza JP, Rowe PJ, Shaw ST Jr. Effect of prior pregnancy and combined oral contraceptives on baseline menstrual blood loss and bleeding response to intrauterine devices. Contraception 1979;20:19-27.

MARCH 1991 VOL. 43 NO. 3

loss.

249