¹Reproductive Biology Unit, Community Medicine Department, National Research Centre.
²Obstetrics and Gynaecology Department, Kasr El-Eini Hospital, Cairo University, Egypt.

Summary

Objective

To assess the impact of the body mass index in resistant PCOD women undergoing laparoscopic ovarian drilling.

Design

Prospective systematically randomized study.

Patients and Methods

Forty infertile anovulatory women who were resistant to clomiphene citrate. All patients had full history taking, physical and gynaecological examination. They were subjected to pelvic ultrasonography and LH, FSH, free testosterone, androstenedione, DHEA-SO₄, SHBG, fasting insulin and progesterone were assayed one month prior to and one month after ovarian drilling.

Results

All women but five women had a BMI of >25 Kg/m². No women with a body mass index less than 25kg/m² had complained of hirsutism. Whilst 2,5 and 5 cases in the BMI range of 25-30, 30-<35 and 35 and more kg/m² had hirsutism respectively. There was a strong positive correlation between BMI, hirsutism, free testosterone and fasting serum insulin levels. It was clear that the more obese the patient the more likely she is hirsute and the more free testosterone and the fasting serum insulin level is. Looking at the possible factors affecting the outcome of laparoscopic ovarian drilling by comparing between the ovulatory (n=36) against the non-ovulatory sub-group (n=4), the BMI was statistically significant P=0.03.

Conclusion

The main indications for laparoscopic ovarian electrocautery treatment in women with PCOD are CC-resistant infertility. To improve results, reliable criteria for patient selection are required. Nevertheless, few data are available predicting the ability to respond favourably to this line of therapy. The response of women to ovarian drilling was related to the BMI and fasting serum insulin level. The ovulation rate in slim and moderately obese women was 98% and dropped to 70% in very obese women.

Key words

PCO, Laparoscopic, electrocautery, BMI and surgical treatment

Introduction

The relationship between fat and reproduction has been recognized for over 20 years. However, the role of obesity in the pathogenesis of PCOD is still unclear. Patients with PCOD are often overweight with a propensity toward abdominal deposition of body fat. Harrington and Balen in their study, found 38% of their PCOD cases to be overweight (Body Mass Index >25kg/m²). Also, obesity was significantly associated with an increased risk of hirsutism, menstrual cycle disturbance and an elevated testosterone concentration. Furthermore, loss of body weight-induced by diet-has important effect, since it reduces blood androgen levels and can improve ovulation ⁽¹⁾. Holte et al studied 49 women with PCOD and 42 controls; both groups covering a wide range of BMI, they found that insulin sensitivity (as measured with the euglycaemic hyperinsulinaemic clamp) showed a linear decline with increasing BMI in both groups much more pronounced in women with PCOD than in control ⁽²⁾.

Aim of the work

The aim of this study was to assess the relation between BMI and PCOD in women undegoing ovarian drilling.

Patients and methods

This was a prospective systematically randomized trial conducted over a two-year period from January 1998 till February 2000. Forty women were recruited into the study. All had chronic anovulation and PCOD as the sole factor of their infertility. The inclusion criteria included women with primary or secondary infertility due to chronic anovulation secondary to resistant PCOD. Resistance was defined as failure of clomiphene citrate in a dose of 100 mg/day for 5 days to induce ovulation for at least six cycles.

PCOD was essentially diagnosed based on detection of an early follicular phase and serum LH/FSH ratio>2. Women above the age of forty or with blocked tube (s) and those cases with male factors or high prolactin level were excluded from the study. Each patient was subjected to a complete history and thorough physical examination. A women was considered infertile should conception failed to occur after at least two years of regular and unprotected intercourse with a fertile husband.

On examination the height in meters and weight in kilograms were recorded Thereafter body mass index was calculated as weight (kg)/height (m)². Patients were segregated into four groups as regards the BMI: a normal group with BMI <25Kg/m² and the group with BMI 25-<30 Kg/m² (overweight). The obese women who have BMI 30- <35Kg/m² while women were considered severely obese if their BMI is more than 35 Kg/m² The clinical evidence of androgen excess (acne or hirsutism) was looked at in areas of hair distribution. Hirsutism was graded into no, mild, moderate and severe according to Raj et al. ⁽³⁾. A mild degree is scored when the dark hair is detected on the upper lip or around the areola and nipple. Moderate hirsutism is assigned should the dark hair is also present on the chest whilst should hair removal is needed at frequent intervals (<3 months) or hirsutism affects the abdomen, back or limbs the severe form is recognized. Breasts were examined for exclusion of clinical evidence of galactorrhoea and thyroid for enlargement. Thereafter abdominal and bimanual pelvic examinations are carried out systematically for any detectable abnormality.

In the month prior to the operative ovarian drilling two blood samples are obtained. One in the early follicular phase (days 3-5) where, a battery of hormones being assessed by RIA procedure. This includes LH, FSH, Free testosterone, androstenedione, DHEA-SO₄, SHBG and fasting serum insulin level. The second sample is collected at mid-luteal phase for progesterone level estimation. Another two samples were collected at the same time as the previous two, one cycle after the operation for assessment of these hormones post-operatively. Data were analyzed with the SPSS-PC (Windows 98 version). Student’s t test was used to compare between two independent means. Chi square test was used to compare between variables with frequencies and logistic multiple regression analysis was used to examine the impact of various factors on the outcome of treatment. Significance was taken as P<0.05.

Results

Forty patients with resistant polycystic ovarian disease underwent laparoscopic ovarian drilling for anovulatory infertility. The mean ±SD of the patients’ age was 29.5±4.9 years. They weigh between 60 to 105 KG with a mean±SD of 78.7±10.7 Kg and their height range between 1.5-1.8 m with a mean±SD of 1.6±0.7 m. The body mass index was in the range of 23.3-39.3 Kg/m² and mean±SD of 30.4±4.9 Kg/m². The mean±SD duration of infertility was 5.4±1.8 years (table 1).

Primary infertility was the presenting symptom in 26 women (65%) while the remaining 14 (35%) had secondary infertility. Menstrual irregularities in the form of oligo-hypo-or oligo-hypomenorrhoea were recorded in 30 patients (75%) whereas 10 (25%) reported regular cycle pattern. No patient had amenorrhoea. All women but 12 (30%) were free from hirsutism. Out of the twelve, 4 (10%) had mild hirsutism while (20%) had moderate affection. No single patient in this study was recorded to have severe hirsutism (table 2 and Figure b).

Upon grouping the forty patients according to their BMI into four groups, five had BMI <25 Kg/m² and 17 were between 25- <30Kg/m². Whereas, in the BMI ranges of 30 -<35 and > 35, 9 patients were found in each group. There was a clear and demonstrable positive association between BMI and free testosterone levels with a higher level the more the woman is obese. Obesity was not only coupled with increasing testosterone measurements but also with a higher fasting insulin level. No women with a body mass index less than 25kg/m² had complained of hirsutism. Whilst 2,5 and 5 cases in the BMI range of 25-<30, >30 and >35kg/m² had excessive hair growth respectively. In a correlation between BMI, hirsutism, free testosterone and fasting serum insulin levels it was clear that the more obese the patient is, the more likely she is hirsute and the more free testosterone and the fasting insulin level (table 3 & 4 and figure c& d).

Following the ovarian drilling 36 women did ovulate and this represents a 90% of the whole patients enrolled in the study. In the responders, the cumulative pregnancy rate after 12-months of follow-up was 43% (17 out of 36). Four cases had a first trimester miscarriage while 13 cases delivered. The overall “take home baby” rate is therefore 36%. Only one case delivered twins.

Looking at the possible factors affecting the outcome of laparoscopic ovarian drilling and by comparing between the ovulatory (n=36) against the non-ovulatory sub-group (n=4) four parameters could be elicited of statistical significance. Firstly was the BMI, where one case in each BMI sub-groups was anovulatory out of 5,17,9 and 9 respectively. This did reach statistical significance P =0.03. Secondly, was the pre-operative mean±SD of fasting serum insulin level that was less in the ovulatory sub-group (38.6±27.4 uIU/ml) than the non-ovulatory sub-group (61.0±19.5 uIU/ml). This was statistically significant P=0.01. Nevertheless, it is the reduction of the post-operative LH value with normalization of LH:FSH ratio and the relevant drop of the fasting insulin level that were of high statistical significance between the ovulatory and the non-ovulatory sub-groups P<0.001 (table 5 & 6).

A striking relation was detected (P=0.0009) in the conception sub-group not with the BMI as in ovulation response but with the type of infertility. Where the patient is more likely to become pregnant should her infertility be of the secondary type (10 pregnant versus 4 non-pregnant). While the reverse is true for primary infertility (7 pregnant versus 15 non-pregnant) following ovarian drilling (table 7).

Discussion

We no longer erroneously think of polycystic ovarian disease in terms of the limited picture of the 1930s as defined by stein and Leventhal of the symptom complex of hyperandrogenism, anovulation and obesity in women with enlarged polycystic ovaries. Now, polycystic ovarian disease is recognized as a heterogeneous condition often with symptoms and signs of elevated androgen levels menstrual irregularity and insulin resistance in women without a well-defined cause of androgen excess. Whilst hyperinsulinaemia has been shown to be an important component of this disease it is not always clinically recognized at an early stage because its symptoms may not be obvious and unless appropriate testing is performed may be missed. Importantly our current understanding indicates that there are links between PCOD and obesity cardiovascular disease and diabetes mellitus^.(4)

From the first description of Stein-Leventhal syndrome clinicians have recognized that at least half of their patients with PCOD are obese. The cause of the obesity associated with PCOD remains unclear, and these women notoriously report that weight loss is difficult with diet and exercise alone. This raises the possibility that these women have a “slow metabolism”. An increasing BMI is correlated with an increased rate of hirsutism, cycle disturbance and infertility. In addition failure to respond to anti-androgens and infertility therapy is much more common in obese than in slim patients.

It is generally assumed that women with hypertrophic fat cells have moderate or sever insulin resistance whereas, those with hyperplastic fat cells have normal insulin receptor function^(5). Obesity is known to depress sex hormone binding globulin levels therefore increasing the proportion of testosterone available for peripheral conversion to oestrone. In addition, there is date that a disproportionately high body fat content causes a rise in extra-ovarian aromatase activity. It is also well established that 20% of obese patients develop hyperinsulinaemia, glucose intolerance and diabetes as a consequence of their body size ⁽¹⁰⁾.

In this study obesity-as defined by BMI> 25Kg/m2 affects 87.5% of the patients. Obesity was clearly associated with increasing insulin resistance that is subsequently reflected in a positive linear manner as higher level of testosterone with increased prevalence of hirsutism. Conway has also reproduced the same result ⁽⁷⁾ The hyperandrogenicity in PCOD women is clearly associated with a preponderance of fat localized to truncal-abdominal sites ⁽⁸⁾. No clear-cut explanation has been put for this preferential truncal adiposity in PCOD women. Although the term male or ‘android’ body fat distribution suggests the involvement of androgen hormones but unluckily testosterone itself seems to promote a reduction in this fat area ⁽⁹⁾. Glucocorticoid receptors are abundant in the abdominal fat tissue and cortisol promotes fat accumulation in the truncal area. Progesterone seems to exert anti-cortisol action through interaction with the cortisol receptor. Therefore, it would seem reasonable to assume that anovulation would result in a more favourable steroid balance for the accumulation of truncal-abdominal fat ⁽¹⁰⁾

Using univariant analysis four factors were significantly and independently influencing the ovulation rate. They were the BMI, pre-operative fasting insulin level the post-operative LH value and normalization of LH: FSH ratio. Thus women who were less obese, less hyperinsulinaemic and ovarian drilling resulted in marked drop of their pre-operative LH concentration had a better chance of ovulation (98%). However by the logistic multivariant analysis it was the BMI that had the best predictive value (96%). The same parameters have been reported before but the serum fasting insulin concentration ⁽¹¹⁾

Conclusion

The main indications for laparoscopic ovarian electrocautery treatment in women with PCOD are CC-resistant infertility. To improve results, reliable criteria for patient selection are required. Nevertheless, few data are available predicting the ability to respond favourably to this line of therapy. In this study, the response of women to ovarian electrocautery was related to the body mass index and fasting serum insulin level. The ovulation rate in slim and moderately obese women was 98% and dropped to 70% in very obese women. However, when ovulation was established the pregnancy rate was independent of body mass index and become dependant mainly on the type of infertility. Women with high serum LH levels responded favourably, whereas those with high levels of testosterone and androstenedione were associated with poor response as regards the pregnancy rate is concerned.

References

1. Harrington D and Balen A (1996): Polycystic ovary syndrome: aetiology and management. B J Hosp Med 56(1): 17-20.

2. Holte J, Bergh T, Berne C and Wide L (1994): Enhanced early insulin response to glucose in relation to insulin resistance in women with polycystic ovary syndrome and normal glucose tolerance. J Clin Endocrinol Metab 78: 1052-1057

3. Raj S, Thompson I, Berger M and Regan L (1978): Clinical aspects of polycystic ovary syndrome. Obstet Gynaecol 49: 552-557.

4. Talbott E, Guzick D and Clerici A (1995): Coronary heart disease risk factors in women with polycystic ovary syndrome. Arteriosclerosis Thrombosis and Vascular Biol 15: 821-826.

5. Bennett S, Todd J and Waterworth D (1997): Association of insulin gene VNTR polymorphism with polycystic ovary syndrome. Lancet 349: 1771-1773.

6. Holte J (1996): Disturbances in insulin secretion and sensitivity in women with the polycystic ovary syndrome. Clin Endocrinol Metab 10: 221-247.

7. Conway G (1996): Polycystic ovary syndrome clinical aspects. Baillieres Clin Endocrinol Metab 10(2): 263-279-

8. Bouchard C, Despres J and Mauriege P (1993): Genetic and non-genetic determinants of regional fat distribution. Endocrine Rev 14: 72-76.

9. Rebuffe-Scrive M, Marin P and Bjorntorp P (1991): Effect of testosterone on abdominal adipose tissue in women. Int J Obesity 15: 791-794.

10. Gennarelli G, Holte J, Wide L, Bernie C and Lithell H (1998): Is there a role for leptin in the endocrine and metabolic aberrations of polycystic ovary syndrome? Hum Repord 13(3): 535-541.

11. Liguori G, Tolino A, Moccia G, Scognamiglio G and Nappi C (1996): Laparoscopic ovarian treatment in infertile patients with polycystic ovarian syndrome: endocrine changes and clinical outcome. Gynaecol Endocrinol 10: 257-264.

Table (1): The physical characteristics of the 40 women who had laparoscopic ovarian drilling for resistant PCOD

The physical Characteristic	Mean ±SD	Range
Age (years)	29.5±4.9	20-39
Weight (Kg)	78.7±10.7	60-105
Height (m)	1.6±0.7	1.5-1.8
BMI (Kg/m²)	30.4±4.9	23.2-39.3
Infertility Duration (years)	5.4±1.8	3-14
Ovarian vol. (ml)	11.1±3.4	5.6-18.3

Figure a: A histogram showing the physical characteristics and the U/S ovarian volume of patients enrolled in this study.

Table (2): The presenting symptoms and the ultrasonographic picture of the 40 women who had laparoscopic ovarian drilling for resistant PCOD

The symptom and Ovarian U/S appearance	Number affected	Percentage
The cycle pattern
Regular	10	25%
Oligomenorrhoea	30	75%
Amenorrhoea	0	0%
Infertility
Primary	26	65%
Secondary	14	35%
Hirsutism
No	28	70%
Mild	4	10%
Moderate	8	20%
Severe	0	0%
Ultrasonography
Typical	28	70%
Atypical	12	30%

Figure b: A figure showing the three hirsute categories seen in our study represented as three slices of a pie.

Table (3): The relationship between free serum testosterone and fasting insulin measurements in infertile CC-resistant PCOD women according to body mass index.

BMI (Kg/m²)	N	Free testosterone (pg/ml)	Fasting insulin (μIU/ml)
<25	5	1.5±0.8	41.6±33.5
25 <30	17	1.6±1.3	42.8±28.8
30 <35	9	1.8±1.3	55.2±23.6
35+	9	2.1±1.2	59.6±19.1

BMI (Kg/m²)

Figure c: A histogram representing the mean serum free testosterone concentration in the 40 patients according to their BMI and its relation to the mean serum fasting insulin levels which is shown as a line.

Table (4): The relationship between hirsutism, free serum testosterone and fasting insulin measurements in infertile CC-resistant PCOD women according to body mass index

BMI (Kg/m²)	Hirsutism	Free testosterone (pg/ml)	Fasting insulin (μIU/ml)
<25	0 of 9
25 <30	2 of 17	1.2±0.5	35.9±24.0
30 <35	5 of 9	2.6±0.6	54.5±13.6
35+	5 of 9	3.2±1.4	63.9±20.8

Figure d: A figure showing the relation between the mean serum free testosterone and fasting insulin concentrations. They were positively correlated with hirsutism.

Table (5): Comparison between the clinical characteristics of ovulatory versus non-ovulatory groups

Physical characteristic		Ovulatory N=36	Non-Ovulatory N=4	P	Sig.
Age (ys)		29.5±4.2	29.0±6.0	0.8	NS
BMI (Kg/m²)		24.2±4.8	31.8±7.0	0.01	Sig.
Duration of infertility (ys)		5.6±1.4	7.1±0.2	0.6	NS
Ov. Vol. (ml)	Pre-	11.2±3.5	10.3±2.6	0.6	NS
	Post-	9.4±2.5	8.6±1.9	0.6	NS

Table (6): The pre-operative hormonal profile (mean & SD) & the post-operative hormonal profile of ovulatory versus non-ovulatory groups

Hormone	Pre-operative Mean ±SD N=40		Post-operative
			Ovulatory N=36		Non-Ovulatory N=4	P	Sig.
FSH (mIU/ml)	4.7 ±1.0	4.7±1.2		5.5±1.2		0.2	NS
LH (mIU/ml)	12.6 ±2.9	4.5±2.6		9.0±1.4		<0.001	HS
LH: FSH ratio	2.7 ±0.8	0.9±0.5		1.6±0.5		<0.001	HS
Androstenedione (pg/ml)	1.6 ±0.7	1.4±0.8		1.3±0.6		0.7	NS
DHEA-SO4 (μg/ml)	132.2 ±72.3	98.0±57.0		74.5±33.0		0.4	NS
Testosterone (pg/ml)	1.7 ±1.2	1.3±1.2		0.6±0.2		0.2	NS
SHBG (nmol/L)	7.9 ±2.5	31.4±13.1		27.1±12.3		0.5	NS
Progesterone (ng/ml)	2.2 ±0.7	22.4±5.5		3.1±0.7		<0.001	HS
Insulin (μIU/ml)	49.9 ±26.8	23.5±21.5		50.7±27.4		0.02	Sig.

Table (7): Comparison between the presenting symptoms, signs and number of drilling holes of conception versus non-conception groups

Physical characteristic	Conception (N=17)	Non-Conception (N=19)	P	Sig.
Cycle pattern
Regular	5	5	0.7	NS
Oligomenorrhoea	12	14
Infertility
Primary	7	15	0.009	Sig.
Secondary	10	4
Hirsutism
No	13	13	0.7	NS
Mild	1	2
Moderate	3	4
BMI (Kg/m²)
<25	2	3	0.8	NS
25 <30	7	8
30 <35	5	4
35+	3	4
U/S picture
Typical	10	17	0.3	NS
Atypical	7	2
Drilling holes
< 8	10	7	0.2	NS
8+	7	12