| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Departments of Clinical Chemistry (S.T., A.R.) and Obstetrics and Gynecology (S.T., H.M., A.-L.H.), University of Oulu and Oulu University Hospital, 90014 Oulu, Finland; Department of Public Health Science and General Practice (S.T., M.-R.J., A.P., U.S.), University of Oulu, 90014 Oulu, Finland; Oulu Regional Institute of Occupational Health (J.L.), 90220 Oulu, Finland; Department of Epidemiology and Public Health (M.-R.J., U.S.), Imperial College Faculty of Medicine, W2 1PG London, United Kingdom; Imperial College Genetics and Genomics Research Institute (M.I.M.), Imperial College Faculty of Medicine, W12 0NN London, United Kingdom; and Department of Reproductive Science and Medicine (S.F.), Institute of Reproductive and Developmental Biology, Imperial College of Science Technology and Medicine, W12 0NN London, United Kingdom
Address all correspondence and requests for reprints to: Saara Taponen, M.D., Department of Clinical Chemistry, University of Oulu, PL 5000, University of Oulu, 90014 Oulu, Finland. E-mail: saara.taponen{at}mail.suomi.net.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Obesity, insulin resistance, and hyperinsulinemia are strongly associated with PCOS (9, 10). Independent of obesity, women with PCOS have cardiovascular risk factors (11, 12). The diagnosis of PCOS in early adulthood would be important to focus on prevention of possible future consequences associated with the syndrome, such as infertility and type 2 diabetes in addition to cardiovascular risk.
The previous studies are mostly based on hospital populations, and to our knowledge no general population-based longitudinal studies on markers or predictors of PCOS exist. We studied a birth cohort of almost 6000 women in Northern Finland (13) to find out the correlation of self-reported symptoms of hirsutism and/or oligomenorrhea with biochemical findings. Our a priori hypothesis was that simple, symptom-based questions are useful to trace women with an endocrine profile typical for PCOS and insulin resistance and thus at risk to develop long-term consequences on health. Because obesity is one of the most important factors that influences clinical and biochemical presentation of women with polycystic ovaries, we stratified the analyses by body mass index categories as well as by weight gain between the ages of 14 and 31 yr to explore their possible modifying effect on natural history of PCOS.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The nested case-control study population derives from a 1966 Northern Finland birth cohort followed since the sixth month of their mother’s pregnancy until the offspring’s age of 31 yr. In 1966, 5964 females were born, 5889 alive. At 14 yr of age, the teenagers answered a postal inquiry. In 1997–1998, 5687 of them were alive and traced for the 31-yr follow-up study when a postal questionnaire was sent to all women. Women still living in the original target or capital city area (n = 4074) were invited for a clinical examination. Of these, 3077 women gave a blood sample and written consent and filled in the postal questionnaire on the symptoms. The questions were: 1) Is your menstrual cycle often more than (over twice a year) 35 d? and 2) Do you have excessive growth of body hair? Of the women who returned the questionnaire (n = 4523), 24% reported symptoms of hirsutism and/or oligomenorrhea (including subjects with amenorrhea), 10.4% reported hirsutism alone, 10.2% reported oligomenorrhea alone, and 3.4% reported both symptoms. Those women who used hormonal contraception (n = 859), were pregnant (n = 211), or used medication for diabetes (n = 13) were excluded from the data. Those who reported symptoms were defined as cases. For each case, two controls were selected randomly from the same birth cohort. The final case-control study included 500 cases and 1026 controls (Fig. 1
).
|
The blood samples were drawn after overnight fasting in the morning (between 0800 and 1100 h). Blood glucose samples were stored in +4 C until analyzed during the same day. Serum insulin samples were stored in -20 C and were analyzed within 7 d of sampling. The T, SHBG, LH, and cortisol samples were stored in -80 C until analyzed. Free androgen index (FAI) was calculated by [testosterone] x 100 divided by [SHBG]. Those who had reported eating, drinking, or smoking during the fasting time (n = 42) were excluded from the statistical analyses of glucose and insulin.
The concentrations of SHBG and LH were analyzed by fluoroimmunoassay (Wallac, Inc. Ltd., Turku, Finland), T by automated chemiluminescence system (ACS-180, Ciba-Corning, Inc., Medfield, MA), cortisol (Orion Diagnostica, Oulunsalo, Finland), and insulin (Pharmacia Diagnostics, Uppsala, Sweden) by RIA, and blood glucose by a glucose dehydrogenase method (Granutest 250, Diagnostica Merck, Darmstadt, Germany). The intra- and interassay coefficients of variation were 1.3% and 5.1% for SHBG, 4.9% and 6.5% for LH, 4.0% and 5.6% for T, 4.0% and 4.3% for cortisol, 5.3% and 7.6% for insulin, and 1.5% and 2.3% for blood glucose. The sensitivity of the insulin assay was 14.35 pmol/liter and of the T assay 0.35 nmol/liter.
Body mass index (BMI) and gain
BMI (kg/m2) was calculated at ages 14 and 31 yr. Overweight and obesity at 14 yr were defined as a BMI at or above the 85th or 95th percentile, respectively. Adult BMI at 31 yr was classified as: normal weight (<25 kg/m2), overweight (25.0–29.9 kg/m2), and obese (
30.0 kg/m2). The subjects were also divided into four subgroups according to their weight development from adolescence to adulthood: 1) normal weight at 14 and 31 yr, 2) normal weight at 14 yr and overweight or obese at 31 yr, 3) overweight or obese at 14 yr and overweight or obese at 31 yr, and 4) overweight or obese at 14 yr and normal weight at 31 yr.
Statistics
Distribution of each outcome variable was skewed to the right both among the cases and controls. Consequently, medians and quartiles are reported and the differences in distributions were tested with the nonparametric Mann-Whitney U test.
Statistical analyses were performed (version 10.1 for Windows; SPSS, Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The hormone levels in the cases, controls, and subjects who reported oligomenorrhea alone (O-group), hirsutism alone (H-group), and both symptoms (OH-group) are presented in Table 1 as medians and lower and upper quartiles. The T, LH, and insulin levels as well as FAI were significantly higher, but SHBG levels alone were lower in the case group than the control group. The only significant difference between the O-group and H-group was observed in LH, which was significantly higher in the O-group. Comparing the H-group to the OH-group or O-group to the OH-group, significantly higher medians of T, LH, FAI, and insulin were observed in the OH-group than H-group or in O-group. The glucose/insulin ratio was significantly lower among the cases than among the controls and lowest in the OH-group. The O- and H-groups did not differ significantly from the whole case group in glucose/insulin ratio. Neither the glucose nor cortisol levels differed significantly among any of the groups.
|
The hormone levels stratified by BMI categories are presented for the cases and controls in Table 2 as medians and lower and upper quartiles. Of the case (control) subjects, 61% (67%) were normal weight, 25% (23%) overweight, and 14% (10%) obese. T and FAI levels were significantly higher but SHBG levels lower among the normal, overweight, and obese cases than among their controls, and they changed in a dose-response manner by increasing BMI category. The difference between the case and control groups remained, independent of BMI. LH levels were significantly higher among the cases than controls only in the group with normal weight. Cortisol levels did not differ between the cases and controls in any of the BMI categories. Insulin and glucose levels increased with increasing BMI, but the cases and controls did not differ from each other in any of the BMI categories. The glucose/insulin ratio decreased with increasing BMI, but there was no statistically significant difference between the cases and controls in any BMI category studied.
|
Figure 2 illustrates T, SHBG, and FAI differences (medians, interquartile ranges) between the cases and controls by weight development from adolescence to adult 31 yr. The subjects were divided into four subgroups according to whether they remained normal weight or overweight or obese, gained overweight or obesity, or lost weight from overweight or obesity to normal weight. The figures indicated that outcome measures change linearly in the first three groups by weight development, but the differences between the cases and controls remain significant except for T in the group with normal weight at 14 yr and overweight or obesity at 31 yr in which the difference shows a similar trend with a higher T level for the case group (P = 0.06). The group with overweight or obesity at 14 yr and normal weight at 31 yr consisted of only 22 cases and 36 controls. There were no statistically significant differences in the hormone levels between the cases and controls in this group, but it is obvious that this group is similar by biochemical findings to the group that remained at normal weight. BMI was highest among the women who had been overweight or obese already at 14 yr [medians 30.5 kg/m2 (cases) and 29.4 kg/m2 (controls)]. In the normal weight at 14 yr and overweight or obese at 31 yr group, the BMI medians were equal, 27.1 kg/m2, for cases and controls. In the normal weight at 14 and 31 yr group the BMI medians were 21.7 kg/m2 (cases) and 21.8 kg/m2 (controls). In the group with overweight or obesity at 14 yr and normal weight at 31 yr, the BMI medians were 23.6 kg/m2 (cases) and 23.2 kg/m2 (controls).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
30 kg/m2) (14, 15, 20). In this study, the 31-yr-old case subjects were defined by their self-reported clinical symptoms, and the proportion of obese women was 14% (10% of the control subjects). The study population base consisted of women born to the 1966 birth cohort who were living in the original target area in Northern Finland or the Helsinki area. This introduces a small hypothetical risk of selection bias, but (e.g. with regard to their circumstances earlier in life) they are well representative of the general female population; as such this is a first of this kind of study in the field with a longitudinal follow-up through to adult life. Because the women who used oral contraceptives or hormonal treatment had to be excluded from the study, it is possible that the individuals with the most severe symptoms and biochemical findings having sought help were left out. This exclusion is, however, likely to dilute the differences between the cases and the controls, which highlights the importance of our findings.
In this study, the median insulin level was significantly higher among the cases than controls. When stratified by BMI categories, the serum insulin levels increased and the glucose/insulin ratio decreased with increasing BMI in both cases and controls, but within the weight categories, insulin levels did not differ significantly between the cases and controls. The glucose/insulin ratio, considered a marker for insulin resistance (21), was, however, significantly lower among the cases than the controls. The levels of SHBG were clearly lower among the cases than the controls throughout all weight categories, which obviously is related to decreased insulin sensitivity and higher insulin secretion in these women (10). It is well documented that insulin secretion is increased, even in lean PCOS subjects (22, 23, 24) and hyperinsulinemia directly reduces serum SHBG levels in PCOS women (25). Hence, SHBG acts as a useful marker of insulin secretion and insulin resistance (26, 27). In accordance with this study, serum SHBG level was shown to be highly effective as a single marker in one epidemiological study for detecting women with PCOS (28). When we compared the cases and controls between the 25% quartile and 75% quartile SHBG concentrations, insulin for the case group was significantly higher in the low SHBG quartile, but T levels did not differ significantly between the quartiles, suggesting that insulin is the major regulator of serum SHBG levels (data not shown).
Body size in adolescence and adulthood was shown to be positively associated with PCOS symptoms in our study population. Holte et al. (5) showed that the negative effects of PCOS and obesity on SHBG levels were independent of each other, whereas T and FAI were affected by obesity only in women with PCOS. In this study, the serum T level seemed to be affected by both PCOS symptoms and obesity. The T levels were higher in case groups throughout all BMI categories, compared with controls, and they increased with increasing BMI in a dose-response manner among both the cases and controls. This effect was also seen in the FAI levels, which also increased with increasing BMI in both the case and control groups, and the difference between symptomatic cases and nonsymptomatic controls was seen in all weight categories.
Weight at 14 yr and its later development predicts the future metabolic changes at the age of 31 yr. In the cases and controls who were normal by weight at 14 and 31 yr, T and FAI were lowest and SHBG highest at 31 yr. The metabolic changes were less favorable in the cases and controls who were normal by weight at the age of 14 and overweight or obese at the age of 31 yr and most unfavorable in those who were overweight or obese both at the age of 14 and 31 yr. In the weight development categories where the subjects remained stable weight or gained overweight or obesity, the cases had higher T and FAI and lower SHBG levels than the controls. The cases and the controls who were overweight or obese at 14 but normal weight at 31 yr had levels of T, SHBG, and FAI similar to the group with normal weight at 14 and 31 yr, suggesting that weight reduction could be beneficial and may prevent unfavorable metabolic changes. The differences in hormone levels between the weight development groups probably are due to, at least partly, the differences in BMI because those cases who were overweight already at the age of 14 yr had the highest BMI.
Elevated serum concentrations of LH are common in women with PCOS (29). In this study there was a significant difference in the serum LH levels between the cases and controls only in the normal weight group, but in overweight and obese women, the difference disappeared. The serum LH levels tended to decrease with increasing weight in the case group but in the control group LH stayed at the same level. This finding is in accord with those of previous studies that have shown an inverse relationship between LH and BMI (5, 30).
In this study, using serum cortisol as an indicator of adrenal steroid production, we could not demonstrate any difference in the adrenal steroidogenesis between the symptomatic cases and controls. However, in the clinically more severe cases of PCOS, cortisol and adrenal androgen secretion is clearly increased (31).
In conclusion, this study shows that self-completed questions concerning oligomenorrhea and hirsutism are useful in detecting women with typical endocrine features of PCOS and signs of decreased insulin sensitivity. Of biochemical markers, SHBG seems to be most useful as an additional screening tool. The unfavorable changes in the endocrine and metabolic parameters, also recorded among the overweight and obese controls, were magnified in the presence of PCOS symptoms, and they may be considered risk factors for health in later life. Special attention, weight reduction and possibly medical treatment, should be paid to the symptomatic overweight or obese women, in whom the most profound changes were observed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BMI, Body mass index; FAI, free androgen index; PCOS, polycystic ovary syndrome; T, testosterone.
Received June 25, 2002.
Accepted October 1, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Koivunen, A. Pouta, S. Franks, H. Martikainen, U. Sovio, A-L. Hartikainen, M.I. McCarthy, A. Ruokonen, A. Bloigu, M-R. Jarvelin, et al. Fecundability and spontaneous abortions in women with self-reported oligo-amenorrhea and/or hirsutism: Northern Finland Birth Cohort 1966 Study Hum. Reprod., September 1, 2008; 23(9): 2134 - 2139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Kumarapeli, R. d. A. Seneviratne, C. N. Wijeyaratne, R. M. S. C. Yapa, and S. H. Dodampahala A Simple Screening Approach for Assessing Community Prevalence and Phenotype of Polycystic Ovary Syndrome in a Semiurban Population in Sri Lanka Am. J. Epidemiol., August 1, 2008; 168(3): 321 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Tzoulaki, M.-R. Jarvelin, A.-L. Hartikainen, M. Leinonen, A. Pouta, M. Paldanius, A. Ruokonen, D. Canoy, U. Sovio, P. Saikku, et al. Size at birth, weight gain over the life course, and low-grade inflammation in young adulthood: northern Finland 1966 birth cohort study Eur. Heart J., April 9, 2008; (2008) ehn105v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Tworoger, A. H. Eliassen, S. A. Missmer, H. Baer, J. Rich-Edwards, K. B. Michels, R. L. Barbieri, M. Dowsett, and S. E. Hankinson Birthweight and Body Size throughout Life in Relation to Sex Hormones and Prolactin Concentrations in Premenopausal Women Cancer Epidemiol. Biomarkers Prev., December 1, 2006; 15(12): 2494 - 2501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sam, R. S. Legro, P. A. Essah, T. Apridonidze, and A. Dunaif Evidence for metabolic and reproductive phenotypes in mothers of women with polycystic ovary syndrome PNAS, May 2, 2006; 103(18): 7030 - 7035. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. L. Powell, L. Haddad, A. Bennett, N. Gharani, U. Sovio, C. J. Groves, K. Rush, M. J. Goh, G. S. Conway, A. Ruokonen, et al. Analysis of Multiple Data Sets Reveals No Association between the Insulin Gene Variable Number Tandem Repeat Element and Polycystic Ovary Syndrome or Related Traits J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 2988 - 2993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Quinkler, B. Sinha, J. W Tomlinson, I. J Bujalska, P. M Stewart, and W. Arlt Androgen generation in adipose tissue in women with simple obesity - a site-specific role for 17{beta}-hydroxysteroid dehydrogenase type 5 J. Endocrinol., November 1, 2004; 183(2): 331 - 342. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Taponen, H. Martikainen, M.-R. Jarvelin, U. Sovio, J. Laitinen, A. Pouta, A.-L. Hartikainen, M. I. McCarthy, S. Franks, M. Paldanius, et al. Metabolic Cardiovascular Disease Risk Factors in Women with Self-Reported Symptoms of Oligomenorrhea and/or Hirsutism: Northern Finland Birth Cohort 1966 Study J. Clin. Endocrinol. Metab., May 1, 2004; 89(5): 2114 - 2118. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Gaasenbeek, B. L. Powell, U. Sovio, L. Haddad, N. Gharani, A. Bennett, C. J. Groves, K. Rush, M. J. Goh, G. S. Conway, et al. Large-Scale Analysis of the Relationship between CYP11A Promoter Variation, Polycystic Ovarian Syndrome, and Serum Testosterone J. Clin. Endocrinol. Metab., May 1, 2004; 89(5): 2408 - 2413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Salmi, H. C. Zisser, and L. Jovanovic Screening for and Treatment of Polycystic Ovary Syndrome in Teenagers Experimental Biology and Medicine, May 1, 2004; 229(5): 369 - 377. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Taponen, S. Ahonkallio, H. Martikainen, R. Koivunen, A. Ruokonen, U. Sovio, A-L. Hartikainen, A. Pouta, J. Laitinen, V. King, et al. Prevalence of polycystic ovaries in women with self-reported symptoms of oligomenorrhoea and/or hirsutism: Northern Finland Birth Cohort 1966 Study Hum. Reprod., May 1, 2004; 19(5): 1083 - 1088. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
