- Email: [email protected]
Carbohydrate metabolism during treatment with estrogen, progestogen, and low-dose oral contraceptives
Carbohydrate metabolism during treatment with estrogen, progestogen, and low-dose oral contraceptives WILLIAM N. SPELLACY, M.D.
Chicago, Illinois The combination oral contraceptives have generally been shown to have an adverse effect on carbohydrate metabolism with resultant elevations of bo1h bk:lod glucose and insulin in users. Studies of the individual steroid components suggest that the estrogen is not at fault. The 19-norprogestins can produce these carbohydrate changel!! and seem to act at the insulin receptor level. Norgestrel, ethynodiol diacetate, and norethindrone alter carbohydrate metabolism, but norgestrel produces the most marked changes. Use of oral contraceptives containing less than 50 ,ug of estrogen resulted in fewer metabolic changes than were seen Wfth the drugs containing higher doses of estrogen. (AM. J. 0BSTET. GYNECOL. 142:732, 1982.)
BoTH A THEORETICAL as well as a practical concern exists about the effects of oral contraceptive steroids (OCs) on carbohydrare metabolism since Waine and associates' first reported, in 1963, a possible adverse effect with OC use. Subsequently, many published reports investigated the effects of pills combining estrogen and progesterone steroids as well as those containing only individual steroids. This article will review some of these data in an attempt to determine if there is evidence of an altered carbohydrate metabolism in OC users, to see if the risk is equal for users of the various steroids, and to evaluate how risk may be minimized by proper drug and patient selection.
Combimltlon oral contraceptives The many published studies dealing with the effects of oral contraceptives on carbohydrate metabolism are difficult to compare because of different research designs that have been used, including cross-sectional versus longitudinal studies, mixed drug types, mixed durations of use, and different blood tests and parameters of measurement. Nevertheless, several reviews of this literature have been done, and in general, they concur that most women who use oral contraceptives From the Department of Obstetrics and Gy!U!colof!Y of The Unit~ersit_v
of Illinois College of Medicine.
Repnnt requests: William N. Spellacy, M.D., Department of Obstetrics and G:vnecolof!Y, The University of Illinois College of Medicine, 840 S. Wood Street, Chicago, lllinoi5 60612.
732
have some elevation of blood glucose and plasma insulin levels and that these changes are most pronounced during a glucose tolerance test.~· a The changes are usually small and somewhat variable. ln one large cross-sectional study in California, the mean rise in blood glucose in users was ll mg/dl. 4 Only about 3% to 5% of women examined develop overt clinical hyperglycemia, and this is usually reversible if the drug is discontinued. 2 A few of this latter group develop a need for insulin treatment that will persist even after the drug is stoppedY· a Since that number is small, it is not clear whether the rate is different from the expected rate for the development of diabetes in a untreated population. 2 The women at greatest risk for developing significant hyperglycemia dl!ring OC use are those considered at risk for developing this condition in general. These include women who have had hyperglycemia in the past (for example, during pregnancy): women with a strong family history of diabetes mellitus: and those who are obese or of older age .2 • '3 Several studies have been done in an attempt to establish how the OC acts on this aspect of the metabolism. Suggestions have included an increase in growth hormone levels in blood: an increase in free cortisol levels in blood; a decrease in tissue pyridoxine levels with a concomitant rise in the anti-insulin tryptophan metabolite xanthurenic acid; and an alteration in the target tissue insulin receptors. This latter area has been explored most often in studies of insulin receptor ac0002-9378/82/060732+03$00.30/0 © 1982 The C. V. Mosby Co.
Carbohydrate metabolism during hormone and OC treatment 733
Volume 142 Number 6, Part 2
150
1(.') ,....
E 01
I
E
I I.)
II
100
I
..::'
I
··"
(.')
1J
- - ISmonth * P<005 ** P
',
' '
**
'e.;,._ '
"'
....E !!
·c:
...........!..
0 0
ai
~~
~
-- --...
60
:l :J
"'
E 0
E
50 Fastmg 0.5
2
3
"'0
50
x---·X 18 month *P<005
**
l':---
"*P
----x***
',
',,
1/, 0
c:
,
'
(.')
__.Control
I ' I ' I ',
:: ,
~
U)
::::l
',,
.,""'*
"'E
~
-control
~~Q
~
~
150
"' E
"' 01 c;oo
I
' ', 'y ..
, ,, , , I
a:
Time in Hours After Glucose
I
Fig. 1. Mean blood glucose values, in milligrams per deciliter, during an oral glucose tolerance test performed before and 18 months after the daily use of 0.25 mg ethynodiol diacetate (!\ = 26).
tivity on circulating blood cells, either red cells or monocytes, done principally because of their accessibility. While it is clear that progesterone can lower in· sulin receptor numbers and affinity, this is not true for all of the progestogens used in the OC.:; It is important, therefore, to review the effects of the individual steroid components of oral contraceptives.
10 0~--+-~--~------~-----,------Fasting 0 5 2 3
Time m Hours After Glucose
Fig. 2. Mean plasma insulin values, in microunits per milliliter, during an oral glucose tolerance test performed before and 18 months after the daily use of0.25 mg ethynodiol diace· tate (N 26).
Table I. Changes in carbohydrate metabolism noted in women who use progestogen alone Progestogen
Estrogens There are few studies on the effects of mestranol and ethinyl estradiol, the two estrogens of the OCs, on carbohydrate metabolism in women, but those that are available fail to show much effect.H This appears to be true at a variety of dose levels. In the large OC studies it was noted that the dosage of the estrogens had little relation to the recorded abnormalities in glucose.
Progestogens Results from the many studies reporting the effects of the individual progestogens in the OC have been variable. While there are studies of injectable progestogens, such as progesterone and medroxyprogesterone acetate (Depo·Provera), these steroids are not used in the OC and are probably not important in this discus· sion. The more commonly used steroids are ethynodiol diacetate, norethindrone, and norgestrel. They are summarized in Table I. Norethindrone. The many published studies of norethindrone have noted some small changes in car· bohydrate metabolism among users. 7 The changes are generally fewer than those seen with the other progestogens studied. however, and are more apparent in the insulin levels than in the glucose levels. Ethynodiol diacetate. There are several studies of this steroid alone, and they show an alteration in both blood glucose and plasma insulin. 8 The results of one
Norethindrone Blood glucose change Plasma insulin change
Mild Moderate
Norgestrel Moderate Moderate
Marked Marked
study of 26 normal women who used 0.25 mg of the steroid are illustrated in Figs. l and 2. In that study a l 00 gm oral glucose tolerance test was done on each woman before and after 18 months of continuous drug use. The mean age of the women was 23.7 ± 1.0 (SEM) years, and mean parity was 2.1 ± 0.3 (SEM). Body weight did not change during the study period [mean control, 142.2 ± 5.1 (SEM) pounds; mean at l8months, 144.0 ± 5.4 (SEM) pounds; t = 1.0173, Pis not significant]. Both blood glucose and immunoassayable plasma insulin levels were measured, and it is clear that both were significantly elevated after drug use. Norgestrel. All of the studies with norgestrel have shown a marked effect on carbohydrate metabolism. 9 Both blood glucose and plasma insulin levels in users are markedly elevated after a few or many cycles. Reports are also available now that demonstrate a sig· nificant decrease in insulin receptors in women taking this steroid. 10
Sub-50 oral contraceptives Because of the concern about adverse metabolic effects that have been shown to accompany OC use, the
734
,\larch l.'i. 198:!
Spellacy
pharmaceutical industry began to formulate new products with a reduced dosage of steroid. The drugs containing less than 50 11-g of estrogen per pill have been termed the .. sub-50" OC products. Only a few studies on their effects on carbohydrate metabolism have been published. Generally, two types of products have been im estigated, both containing ethinyl estradiol at 30 to ~15 11-g per pill. The two product groups were divided on the basis of their progestogen type as to those containing norgestrel and those containing norethindrone .11 These studies show that the products containing norgestrel are associated with an alteration of glucose and/or insulin levels, whereas those containing norethindrone are not associated with any adverse carbohydrate changes.'' These data support the belief that the progestogen steroid is important in carbohydrate metabolic changes and further emphasize the potency of norgestrel in this regard. Further long-term studies with large numbers of women who use the sub-50 OC with norethindrone will be necessary in order to confirm the assumption that these preparations do not seem to alter carbohydrate metabolism.
Comment Family planning specialists have voiced a growing concern that the adverse metabolic effects associated with oral contraceptives could be a potential long-range problem for the user in terms of clinical difficulties, such as advanced atherogenesis and premature \'ascular disease. These metabolic effects have included changes in blood pressure, carbohydrate metabolism, and lipid metabolism. In order to reduce these risks we need to better understand the steroids responsible for the changes and the mechanisms involved. The studies of carbohydrate metabolism with con-
REFERENCES I. Waine, H .. Frieden, E. H., Caplan, H. 1., et al.: Meta-
bolic effects of Enovid in rheumatoid patients, Arthritis Rheum. 6:796, 1963. 2. Kalkhoff, R. K.: Effects of oral contraceptive agents on carbohydrate metabolism, J. Steroid Biochem. 6:949,
1975. 3. Spella('}. W. ~.:Carbohydrate metabolism in male infertility and female fertility-control patients, Fertil. Steril.
27:1132, 1976. .t. Phillips, N .. and Duffy, T.: One-hour glucose tolerance in relationship to the use of contraceptive drugs, AM. J. 0BSTET. GYNECOI-. ll6:91, 1973. 5. Tsibris,J. C. M., Raynor, L. 0., Buhi, W. C., et al.: Insulin receptors in circulating erythrocytes and monocytes from women on oral contraceptives or pregnant women near term,]. Clin. Endocrinol. 51:711, 1980. 6. Spellacy. W. N., Buhi, W. C., and Birk, S. A.: The effect of estrogens on carbohydrate metabolism: Glucose, insulin and growth hormone studies on one hundred and
Am, J. Obstet. Gmecol.
traceptive steroids suggest that the synthetic estrogem. namely, mestranol and ethinyl estradiol, produce few adverse effects. Indeed, animal studie~ as well as in vitro membrane insulin receptor studies indicate that at certain dosages the estrogens improve carbohydrate metabolism. ;\; umerous studies on the ,·arious progestogens ha\·e shown. in general, that the 19-norprogestogens adversely affect this metabolism The reports also demonstrate different potencies. with norgestrd having the greatest and norethindrone the least dkct.. These data are further supported bv insulin receptor stndies that show major changes with norgestrel and minimal changes with norethindrone. Studies with the combination low-dose oral contraceptives also agree: The 30 to 35 11-g estrogen products with norgestrel cause some carbohydrate changes, whereas those with norethindrone are relatively free of effects. \Vhile these short-term metabolic -;unT\S are encouraging, they do not demonstrate that blood vessels will be either unaffected or protected hv use of the newer products. Only long-term imestigations will make such clinical data a\·ailable. It does seem prudent, however, in the interim to begin to apply the results clinically. At present it does not seem appropriate to begin giving a patient an OC with more than 50 p.g of estrogen. The occasional patient will have to be given a product with more than 50 JLg of estrogen, but she \\·ill not be common. The type and dose of progestins will also be important. Products with weak progestins, such as norethindrone, administered at a Jm,· dosage are desirable. As \\'omen receive a more complete pre-OC medical eYaluation and periodic screening and as proper drug selection is used, the data. we hope. will show that the risk of OC use is far less than the benefit.
seventy-one women ingesting Premarin, mestranol, and ethinyl estradiol for six months, AM. J. 0BSTET. GYNECOL.
ll4:378, 1972. 7. Spellacy, W. N., Buhi, W. C., and Birk. S. A.: Effects of norethindrone on carbohydrate and lipid metabolism, Obstet. Gynecol. 46:560, 1975. 8. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: Carbohydrate and lipid metabolic studies before and after one year of treatment with ethynodiol diacetate in "normal" women, Fertil. Steril. 27:900, 1976. 9. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: The effects of norgestrel on carbohydrate and lipid metabolism over one year, AM.j. 0BSTET. GYNECOL. 125:984, 1976. 10. DePirro, R., Forte, F., Bertoli, A., et al.: Changes in insulin receptors during oral contraception, J. Clin. Endocrinol. 52:29, 1981. ll. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: Carbohydrate metabolism prospectively studied in women using a low-estrogen oral contraceptive for six months, Contraception 20:137, 1979.
Chicago, Illinois The combination oral contraceptives have generally been shown to have an adverse effect on carbohydrate metabolism with resultant elevations of bo1h bk:lod glucose and insulin in users. Studies of the individual steroid components suggest that the estrogen is not at fault. The 19-norprogestins can produce these carbohydrate changel!! and seem to act at the insulin receptor level. Norgestrel, ethynodiol diacetate, and norethindrone alter carbohydrate metabolism, but norgestrel produces the most marked changes. Use of oral contraceptives containing less than 50 ,ug of estrogen resulted in fewer metabolic changes than were seen Wfth the drugs containing higher doses of estrogen. (AM. J. 0BSTET. GYNECOL. 142:732, 1982.)
BoTH A THEORETICAL as well as a practical concern exists about the effects of oral contraceptive steroids (OCs) on carbohydrare metabolism since Waine and associates' first reported, in 1963, a possible adverse effect with OC use. Subsequently, many published reports investigated the effects of pills combining estrogen and progesterone steroids as well as those containing only individual steroids. This article will review some of these data in an attempt to determine if there is evidence of an altered carbohydrate metabolism in OC users, to see if the risk is equal for users of the various steroids, and to evaluate how risk may be minimized by proper drug and patient selection.
Combimltlon oral contraceptives The many published studies dealing with the effects of oral contraceptives on carbohydrate metabolism are difficult to compare because of different research designs that have been used, including cross-sectional versus longitudinal studies, mixed drug types, mixed durations of use, and different blood tests and parameters of measurement. Nevertheless, several reviews of this literature have been done, and in general, they concur that most women who use oral contraceptives From the Department of Obstetrics and Gy!U!colof!Y of The Unit~ersit_v
of Illinois College of Medicine.
Repnnt requests: William N. Spellacy, M.D., Department of Obstetrics and G:vnecolof!Y, The University of Illinois College of Medicine, 840 S. Wood Street, Chicago, lllinoi5 60612.
732
have some elevation of blood glucose and plasma insulin levels and that these changes are most pronounced during a glucose tolerance test.~· a The changes are usually small and somewhat variable. ln one large cross-sectional study in California, the mean rise in blood glucose in users was ll mg/dl. 4 Only about 3% to 5% of women examined develop overt clinical hyperglycemia, and this is usually reversible if the drug is discontinued. 2 A few of this latter group develop a need for insulin treatment that will persist even after the drug is stoppedY· a Since that number is small, it is not clear whether the rate is different from the expected rate for the development of diabetes in a untreated population. 2 The women at greatest risk for developing significant hyperglycemia dl!ring OC use are those considered at risk for developing this condition in general. These include women who have had hyperglycemia in the past (for example, during pregnancy): women with a strong family history of diabetes mellitus: and those who are obese or of older age .2 • '3 Several studies have been done in an attempt to establish how the OC acts on this aspect of the metabolism. Suggestions have included an increase in growth hormone levels in blood: an increase in free cortisol levels in blood; a decrease in tissue pyridoxine levels with a concomitant rise in the anti-insulin tryptophan metabolite xanthurenic acid; and an alteration in the target tissue insulin receptors. This latter area has been explored most often in studies of insulin receptor ac0002-9378/82/060732+03$00.30/0 © 1982 The C. V. Mosby Co.
Carbohydrate metabolism during hormone and OC treatment 733
Volume 142 Number 6, Part 2
150
1(.') ,....
E 01
I
E
I I.)
II
100
I
..::'
I
··"
(.')
1J
- - ISmonth * P<005 ** P
',
' '
**
'e.;,._ '
"'
....E !!
·c:
...........!..
0 0
ai
~~
~
-- --...
60
:l :J
"'
E 0
E
50 Fastmg 0.5
2
3
"'0
50
x---·X 18 month *P<005
**
l':---
"*P
----x***
',
',,
1/, 0
c:
,
'
(.')
__.Control
I ' I ' I ',
:: ,
~
U)
::::l
',,
.,""'*
"'E
~
-control
~~Q
~
~
150
"' E
"' 01 c;oo
I
' ', 'y ..
, ,, , , I
a:
Time in Hours After Glucose
I
Fig. 1. Mean blood glucose values, in milligrams per deciliter, during an oral glucose tolerance test performed before and 18 months after the daily use of 0.25 mg ethynodiol diacetate (!\ = 26).
tivity on circulating blood cells, either red cells or monocytes, done principally because of their accessibility. While it is clear that progesterone can lower in· sulin receptor numbers and affinity, this is not true for all of the progestogens used in the OC.:; It is important, therefore, to review the effects of the individual steroid components of oral contraceptives.
10 0~--+-~--~------~-----,------Fasting 0 5 2 3
Time m Hours After Glucose
Fig. 2. Mean plasma insulin values, in microunits per milliliter, during an oral glucose tolerance test performed before and 18 months after the daily use of0.25 mg ethynodiol diace· tate (N 26).
Table I. Changes in carbohydrate metabolism noted in women who use progestogen alone Progestogen
Estrogens There are few studies on the effects of mestranol and ethinyl estradiol, the two estrogens of the OCs, on carbohydrate metabolism in women, but those that are available fail to show much effect.H This appears to be true at a variety of dose levels. In the large OC studies it was noted that the dosage of the estrogens had little relation to the recorded abnormalities in glucose.
Progestogens Results from the many studies reporting the effects of the individual progestogens in the OC have been variable. While there are studies of injectable progestogens, such as progesterone and medroxyprogesterone acetate (Depo·Provera), these steroids are not used in the OC and are probably not important in this discus· sion. The more commonly used steroids are ethynodiol diacetate, norethindrone, and norgestrel. They are summarized in Table I. Norethindrone. The many published studies of norethindrone have noted some small changes in car· bohydrate metabolism among users. 7 The changes are generally fewer than those seen with the other progestogens studied. however, and are more apparent in the insulin levels than in the glucose levels. Ethynodiol diacetate. There are several studies of this steroid alone, and they show an alteration in both blood glucose and plasma insulin. 8 The results of one
Norethindrone Blood glucose change Plasma insulin change
Mild Moderate
Norgestrel Moderate Moderate
Marked Marked
study of 26 normal women who used 0.25 mg of the steroid are illustrated in Figs. l and 2. In that study a l 00 gm oral glucose tolerance test was done on each woman before and after 18 months of continuous drug use. The mean age of the women was 23.7 ± 1.0 (SEM) years, and mean parity was 2.1 ± 0.3 (SEM). Body weight did not change during the study period [mean control, 142.2 ± 5.1 (SEM) pounds; mean at l8months, 144.0 ± 5.4 (SEM) pounds; t = 1.0173, Pis not significant]. Both blood glucose and immunoassayable plasma insulin levels were measured, and it is clear that both were significantly elevated after drug use. Norgestrel. All of the studies with norgestrel have shown a marked effect on carbohydrate metabolism. 9 Both blood glucose and plasma insulin levels in users are markedly elevated after a few or many cycles. Reports are also available now that demonstrate a sig· nificant decrease in insulin receptors in women taking this steroid. 10
Sub-50 oral contraceptives Because of the concern about adverse metabolic effects that have been shown to accompany OC use, the
734
,\larch l.'i. 198:!
Spellacy
pharmaceutical industry began to formulate new products with a reduced dosage of steroid. The drugs containing less than 50 11-g of estrogen per pill have been termed the .. sub-50" OC products. Only a few studies on their effects on carbohydrate metabolism have been published. Generally, two types of products have been im estigated, both containing ethinyl estradiol at 30 to ~15 11-g per pill. The two product groups were divided on the basis of their progestogen type as to those containing norgestrel and those containing norethindrone .11 These studies show that the products containing norgestrel are associated with an alteration of glucose and/or insulin levels, whereas those containing norethindrone are not associated with any adverse carbohydrate changes.'' These data support the belief that the progestogen steroid is important in carbohydrate metabolic changes and further emphasize the potency of norgestrel in this regard. Further long-term studies with large numbers of women who use the sub-50 OC with norethindrone will be necessary in order to confirm the assumption that these preparations do not seem to alter carbohydrate metabolism.
Comment Family planning specialists have voiced a growing concern that the adverse metabolic effects associated with oral contraceptives could be a potential long-range problem for the user in terms of clinical difficulties, such as advanced atherogenesis and premature \'ascular disease. These metabolic effects have included changes in blood pressure, carbohydrate metabolism, and lipid metabolism. In order to reduce these risks we need to better understand the steroids responsible for the changes and the mechanisms involved. The studies of carbohydrate metabolism with con-
REFERENCES I. Waine, H .. Frieden, E. H., Caplan, H. 1., et al.: Meta-
bolic effects of Enovid in rheumatoid patients, Arthritis Rheum. 6:796, 1963. 2. Kalkhoff, R. K.: Effects of oral contraceptive agents on carbohydrate metabolism, J. Steroid Biochem. 6:949,
1975. 3. Spella('}. W. ~.:Carbohydrate metabolism in male infertility and female fertility-control patients, Fertil. Steril.
27:1132, 1976. .t. Phillips, N .. and Duffy, T.: One-hour glucose tolerance in relationship to the use of contraceptive drugs, AM. J. 0BSTET. GYNECOI-. ll6:91, 1973. 5. Tsibris,J. C. M., Raynor, L. 0., Buhi, W. C., et al.: Insulin receptors in circulating erythrocytes and monocytes from women on oral contraceptives or pregnant women near term,]. Clin. Endocrinol. 51:711, 1980. 6. Spellacy. W. N., Buhi, W. C., and Birk, S. A.: The effect of estrogens on carbohydrate metabolism: Glucose, insulin and growth hormone studies on one hundred and
Am, J. Obstet. Gmecol.
traceptive steroids suggest that the synthetic estrogem. namely, mestranol and ethinyl estradiol, produce few adverse effects. Indeed, animal studie~ as well as in vitro membrane insulin receptor studies indicate that at certain dosages the estrogens improve carbohydrate metabolism. ;\; umerous studies on the ,·arious progestogens ha\·e shown. in general, that the 19-norprogestogens adversely affect this metabolism The reports also demonstrate different potencies. with norgestrd having the greatest and norethindrone the least dkct.. These data are further supported bv insulin receptor stndies that show major changes with norgestrel and minimal changes with norethindrone. Studies with the combination low-dose oral contraceptives also agree: The 30 to 35 11-g estrogen products with norgestrel cause some carbohydrate changes, whereas those with norethindrone are relatively free of effects. \Vhile these short-term metabolic -;unT\S are encouraging, they do not demonstrate that blood vessels will be either unaffected or protected hv use of the newer products. Only long-term imestigations will make such clinical data a\·ailable. It does seem prudent, however, in the interim to begin to apply the results clinically. At present it does not seem appropriate to begin giving a patient an OC with more than 50 p.g of estrogen. The occasional patient will have to be given a product with more than 50 JLg of estrogen, but she \\·ill not be common. The type and dose of progestins will also be important. Products with weak progestins, such as norethindrone, administered at a Jm,· dosage are desirable. As \\'omen receive a more complete pre-OC medical eYaluation and periodic screening and as proper drug selection is used, the data. we hope. will show that the risk of OC use is far less than the benefit.
seventy-one women ingesting Premarin, mestranol, and ethinyl estradiol for six months, AM. J. 0BSTET. GYNECOL.
ll4:378, 1972. 7. Spellacy, W. N., Buhi, W. C., and Birk. S. A.: Effects of norethindrone on carbohydrate and lipid metabolism, Obstet. Gynecol. 46:560, 1975. 8. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: Carbohydrate and lipid metabolic studies before and after one year of treatment with ethynodiol diacetate in "normal" women, Fertil. Steril. 27:900, 1976. 9. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: The effects of norgestrel on carbohydrate and lipid metabolism over one year, AM.j. 0BSTET. GYNECOL. 125:984, 1976. 10. DePirro, R., Forte, F., Bertoli, A., et al.: Changes in insulin receptors during oral contraception, J. Clin. Endocrinol. 52:29, 1981. ll. Spellacy, W. N., Buhi, W. C., and Birk, S. A.: Carbohydrate metabolism prospectively studied in women using a low-estrogen oral contraceptive for six months, Contraception 20:137, 1979.