373 assessed. We would be grateful to receive other patients with pineal tumours.
serum
samples from
Department of Medicine, General Hospital, Birmingham B4 6NH
S. G. BARBER
University of Bradford
J. A. SMITH
Dudley Road Hospital, Birmingham
D. H. COVE
Women’s
S. C. H. SMITH
Hospital, Birmingham
Queen Elizabeth Hospital, Birmingham
D. R. LONDON
SMOKING, OBESITY, AND THE MENOPAUSE SiR,—Two reports in The Lancet12 have confirmed
among
non-smoking women over age 50. Among the 500 studied, hysterectomy had been done, for each decade age 50, with similar frequency among non-smokers and
women
my
observations3 and those of Hammond4 that habitual cigarette
smoking is associated with an earlier natural menopause. Mattison and Thorgeirsson5 have suggested that this change may be due to chemicals absorbed from tobacco smoke, deposited in ovarian tissue and there modified by ovarian enzymes into agents known to be carcinogenic. I suggest that a different mechanism may contribute to the relation between smoking and early menopause-namely, the relative lack of obesity among smokers.
Non-smokers are more likely than smokers to be obese6-8 and obese women tend to have a later menopause.9 To determine whether early menopause was independently related to both smoking habits and lack of obesity, I analysed data for 500 consecutive women aged 40-69 interviewed in my general internal medicine practice. 236 had had a natural menopause. The other 264 women were excluded because they were premenopausal, had had ovarian or uterine surgery, or had received supplemental hormones for treatment of menopausal symptoms, thereby obscuring the onset of their natural menopause. Weights were compared with those recommended for height in standard life-insurance tables for women of medium build. A woman with a smoking history of 5 pack years before age 45 was defined as a cigarette smoker. Smoking and obesity were independently associated with age at menopause (see table). Some of the difference between the average menopausal age of smokers and non-smokers will be due to the greater tendency of non-smokers to be obese and of smokers to be underweight. In addition, however, significant differences between smokers and non-smokers similarly obese persisted, an observation compatible with the hypothesis of Mattison and Thorgeirsson. It seems likely that the delayed menopauseof obese women is related to the production of oestrogen by adipose tissue which can transform an adrenal steroid (androstenedione) with little oestrogen activity into oestrone, a potent cestrogen.10 11 The much greater production of oestrone by obese menopausal women has been offered as an explanation for the greatly prolonged presence, among them, of cestrogenic effect in vaginal smears and urinary sediment12 13 as well as their inherent protection against osteoporosis.3 This seems likely to explain the delayed menopause of obese women as well. Of additional interest is the high frequency of hysterectomy Jick, H., Porter, J., Morrison, A. S. Lancet, 1977, i, 1354. Bailey, A., Robinson, D., Vessey, M. ibid. 1977, ii, 722. Daniell, H. W. Archs intern. Med. 1976, 136, 298. 4. Hammond, E.C. Archs envir. Hlth, 1961, 3, 146. 5. Mattison, D. R., Thorgeirsson, Lancet, 1978, i, 187. 6. Higgins, M. W., Kjelsberg, M. Am. J. Epidem. 1967, 86, 60. 7. Khosla, T., Lowe, C. R. Br. med. J. 1971, iv, 10. 8. Comstock, G. W., Stone, R.W. Archs envir. Hlth. 1972, 24, 271. 9. MacMahon, B., Worcester, J. Age at Menopause: Natn. Center Hlth Stat., series 11, no. 19. Washington, D. C., 1966. 10. Schindler, A. E., Ebert, A., Friedrichs, E.J. clin. Endocr. 1972, 35, 627. 11. Grodin, J. M., Siiteri, P. K., MacDonald, P. C. J. clin. Endocr. Metab. 1973, 36, 207. 12. DeWaard F., Oettle, A. G. Cancer. 1965, 18, 450. 13. DeWard, F., Schwarz, F. Acta Cytol. 1964, 8, 449. 1. 2. 3.
AVERAGE AGE AT NATURAL MENOPAUSE ANALYSED BY HISTORY OF CIGARETTE SMOKING AND PRESENCE OF OBESITY
before smokers and for obese and non-obese. After age 50, however, while 40% of women were smokers, all but 3 of the 29 who had a hysterectomy were non-smokers, and most of these were over-weight. Most of these operations had been done because of erratic or excessive menstrual bleeding which, seems likely to reflect prolonged oestrogen effect among these women. 2020 Court Street, Redding, California 96001, U.S.A.
HARRY W. DANIELL
SCREENING FOR NEURAL-TUBE DEFECTS
SIR,-In their comments on our article on voluntary maternal serum-alpha-fetoprotein (A.F.P.) screening for fetal neuraltube defects, Dr Wald and colleagues (July 22, p. 216) point out that there are no material inconsistencies between our study and the U.K. collaborative study.1 Ours was the largest single group of patients with neural-tube defects contributed to the U.K. study and we certainly did not intend to imply that there were important differences of fact. The major difference relates rather to screening policy. However, we would like to clarify a misunderstanding about the classification of closed neural-tube defects which tends to give a misleadingly low estimate of the screening sensitivity for all spina-bifida pregnancies in the West of Scotland study. The 14 closed neural-tube defects encountered in phase n of our study can all be unambiguously defined as closed lesions because they were covered at birth either by full-thickness skin or by opaque fibrous tissue and keratinising squamous epithelium. There were 2 iniencephalics with closed posterior spina bifida and hydrocephalus (one of which had to be decompressed to allow delivery), 5 meningoceles, 1 encephalocele, 2 myeloceles (one of which was ruptured during delivery), and 4 fetuses with spina-bifida occulta each identified at birth because of an overlying hairy noevus. Although we included the latter 4 with the other closed defects, they would not normally be classified as spina-bifida cystica malformations and should perhaps have been excluded. If they are excluded, the sensitivity of our test for all spina-bifida pregnancies in phase n is 14 of 26 (53.8%) and not, as Wald suggests, 14 of 30 (47%). What is clear, however, is that none of the closed lesions in our series could have conceivably been classified as open and missed, so that the 81.2% sensitivity quoted for open spina bifida is a true estimate which we believe compares favourably with the 64% sensitivity estimated from the U.K. study by Wald and colleagues. Admittedly, the numbers in both series are small and the confidence limits wide, and it remains to be seen which will prove the more reliable estimate. We feel that our study may have an advantage in that particular attention has been paid to the accurate estimation of gestation by obstetric ultrasound in both patients and controls, which may not have been equally available to the other 18 centres contributing to the U.K. study. 1. U.K. Collaborative
Study
on
Alpha-fetoprotein
Defects. Lancet, 1977, i, 1323.
in
Relation
to
Neural-tube
serum
samples from
Department of Medicine, General Hospital, Birmingham B4 6NH
S. G. BARBER
University of Bradford
J. A. SMITH
Dudley Road Hospital, Birmingham
D. H. COVE
Women’s
S. C. H. SMITH
Hospital, Birmingham
Queen Elizabeth Hospital, Birmingham
D. R. LONDON
SMOKING, OBESITY, AND THE MENOPAUSE SiR,—Two reports in The Lancet12 have confirmed
among
non-smoking women over age 50. Among the 500 studied, hysterectomy had been done, for each decade age 50, with similar frequency among non-smokers and
women
my
observations3 and those of Hammond4 that habitual cigarette
smoking is associated with an earlier natural menopause. Mattison and Thorgeirsson5 have suggested that this change may be due to chemicals absorbed from tobacco smoke, deposited in ovarian tissue and there modified by ovarian enzymes into agents known to be carcinogenic. I suggest that a different mechanism may contribute to the relation between smoking and early menopause-namely, the relative lack of obesity among smokers.
Non-smokers are more likely than smokers to be obese6-8 and obese women tend to have a later menopause.9 To determine whether early menopause was independently related to both smoking habits and lack of obesity, I analysed data for 500 consecutive women aged 40-69 interviewed in my general internal medicine practice. 236 had had a natural menopause. The other 264 women were excluded because they were premenopausal, had had ovarian or uterine surgery, or had received supplemental hormones for treatment of menopausal symptoms, thereby obscuring the onset of their natural menopause. Weights were compared with those recommended for height in standard life-insurance tables for women of medium build. A woman with a smoking history of 5 pack years before age 45 was defined as a cigarette smoker. Smoking and obesity were independently associated with age at menopause (see table). Some of the difference between the average menopausal age of smokers and non-smokers will be due to the greater tendency of non-smokers to be obese and of smokers to be underweight. In addition, however, significant differences between smokers and non-smokers similarly obese persisted, an observation compatible with the hypothesis of Mattison and Thorgeirsson. It seems likely that the delayed menopauseof obese women is related to the production of oestrogen by adipose tissue which can transform an adrenal steroid (androstenedione) with little oestrogen activity into oestrone, a potent cestrogen.10 11 The much greater production of oestrone by obese menopausal women has been offered as an explanation for the greatly prolonged presence, among them, of cestrogenic effect in vaginal smears and urinary sediment12 13 as well as their inherent protection against osteoporosis.3 This seems likely to explain the delayed menopause of obese women as well. Of additional interest is the high frequency of hysterectomy Jick, H., Porter, J., Morrison, A. S. Lancet, 1977, i, 1354. Bailey, A., Robinson, D., Vessey, M. ibid. 1977, ii, 722. Daniell, H. W. Archs intern. Med. 1976, 136, 298. 4. Hammond, E.C. Archs envir. Hlth, 1961, 3, 146. 5. Mattison, D. R., Thorgeirsson, Lancet, 1978, i, 187. 6. Higgins, M. W., Kjelsberg, M. Am. J. Epidem. 1967, 86, 60. 7. Khosla, T., Lowe, C. R. Br. med. J. 1971, iv, 10. 8. Comstock, G. W., Stone, R.W. Archs envir. Hlth. 1972, 24, 271. 9. MacMahon, B., Worcester, J. Age at Menopause: Natn. Center Hlth Stat., series 11, no. 19. Washington, D. C., 1966. 10. Schindler, A. E., Ebert, A., Friedrichs, E.J. clin. Endocr. 1972, 35, 627. 11. Grodin, J. M., Siiteri, P. K., MacDonald, P. C. J. clin. Endocr. Metab. 1973, 36, 207. 12. DeWaard F., Oettle, A. G. Cancer. 1965, 18, 450. 13. DeWard, F., Schwarz, F. Acta Cytol. 1964, 8, 449. 1. 2. 3.
AVERAGE AGE AT NATURAL MENOPAUSE ANALYSED BY HISTORY OF CIGARETTE SMOKING AND PRESENCE OF OBESITY
before smokers and for obese and non-obese. After age 50, however, while 40% of women were smokers, all but 3 of the 29 who had a hysterectomy were non-smokers, and most of these were over-weight. Most of these operations had been done because of erratic or excessive menstrual bleeding which, seems likely to reflect prolonged oestrogen effect among these women. 2020 Court Street, Redding, California 96001, U.S.A.
HARRY W. DANIELL
SCREENING FOR NEURAL-TUBE DEFECTS
SIR,-In their comments on our article on voluntary maternal serum-alpha-fetoprotein (A.F.P.) screening for fetal neuraltube defects, Dr Wald and colleagues (July 22, p. 216) point out that there are no material inconsistencies between our study and the U.K. collaborative study.1 Ours was the largest single group of patients with neural-tube defects contributed to the U.K. study and we certainly did not intend to imply that there were important differences of fact. The major difference relates rather to screening policy. However, we would like to clarify a misunderstanding about the classification of closed neural-tube defects which tends to give a misleadingly low estimate of the screening sensitivity for all spina-bifida pregnancies in the West of Scotland study. The 14 closed neural-tube defects encountered in phase n of our study can all be unambiguously defined as closed lesions because they were covered at birth either by full-thickness skin or by opaque fibrous tissue and keratinising squamous epithelium. There were 2 iniencephalics with closed posterior spina bifida and hydrocephalus (one of which had to be decompressed to allow delivery), 5 meningoceles, 1 encephalocele, 2 myeloceles (one of which was ruptured during delivery), and 4 fetuses with spina-bifida occulta each identified at birth because of an overlying hairy noevus. Although we included the latter 4 with the other closed defects, they would not normally be classified as spina-bifida cystica malformations and should perhaps have been excluded. If they are excluded, the sensitivity of our test for all spina-bifida pregnancies in phase n is 14 of 26 (53.8%) and not, as Wald suggests, 14 of 30 (47%). What is clear, however, is that none of the closed lesions in our series could have conceivably been classified as open and missed, so that the 81.2% sensitivity quoted for open spina bifida is a true estimate which we believe compares favourably with the 64% sensitivity estimated from the U.K. study by Wald and colleagues. Admittedly, the numbers in both series are small and the confidence limits wide, and it remains to be seen which will prove the more reliable estimate. We feel that our study may have an advantage in that particular attention has been paid to the accurate estimation of gestation by obstetric ultrasound in both patients and controls, which may not have been equally available to the other 18 centres contributing to the U.K. study. 1. U.K. Collaborative
Study
on
Alpha-fetoprotein
Defects. Lancet, 1977, i, 1323.
in
Relation
to
Neural-tube
