Risk factors for epithelial ovarian cancer in the female population of Belgrade , Serbia : A case-control study

Background/Aim. Ovarian cancer (OC) comprises 3% of all cancers, but it is the fifth most common cause of cancer death in women. The aim of this case-control study was to determine the risk factors for OC in the female population of Belgrade, Serbia. Methods. A total of 80 consecutive patients were enrolled in the study between 2006 and 2008 in two national referral centers for OC in Serbia. The control subjects were recruited during the regular gynecological check-ups in the Public Health Center of the corresponding municipalities. All the study participants were interviewed during their visits to the above mentioned institutions by two physicians using the same questionnaire. In order to analyze the influence of specific exposure to the risk of the disease, we categorized variables according to the cut-off values. Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated separately for each variable using univariate conditional logistic regression analysis. Results. There were no statistically significant differences in educational level, years of schooling, occupational and employment status between patients with OC and women in the control group. Oral contraceptives use and other contraceptive methods (condoms, mechanical contraceptive devices) were highly statistically significantly more frequent among women in the control group (OR = 0.2, 95% CI 0.1–0.7, p = 0.005; OR = 0.1, 95% CI 0.01–0.5, p = 0.001, respectively). The patients with OC practiced sports for 6.3 ± 2.1 years, and controls for 11.8 ± 9.9 years. Sport and recreation activities were statistically significantly protective (OR = 0.2, p = 0.011; OR = 0.4, p = 0.019). Tea consumption on daily basis had a highly statistically significat protective effect (OR = 0.3, p = 0.001). Conclusions. Oral contraceptives use and physical activity were independent protective factors for OC in this study.


Introduction
Ovarian cancer (OC) comprises 3% of all cancers, but it is the fifth most common cause of cancer death in women 1 .The 5-year relative survival rate ranges from 30% to 45%, without significant improvement in the past years even though the new methods in therapy have been used [2][3][4] .The highest incidence rates of OC have been registered in Scandinavia, Eastern Europe and Canada, where it varies between 10 and 15 per 100,000 women 5 .The lowest ones were found in Asia (excluding Japan) and Africa, with less than 5 per 100,000 women 6 .In Serbia, the standardized incidence rate in 1999 was 9.7 per 100.000, while in 2005 it rose up to 11.5 per 100.000 7 .
Risk factors for OC are still not well established.Age, family history of OC, infertility treatment and assisted fertilization, hormonal substitution in menopause, and obesity are potential factors in favor of developing the OC 8,9 .Also, it has been noted that nulliparous women have an increased risk for OC 8,10 .On the other hand, it has been well defined and quantified that the use of oral contraceptives decreases the risk 6 as well as multiple pregnancies.In addition, tubal ligation, hysterectomy and lactation are found to be protective factors, too 11 .
Genetic factors also play an important role in the etiology of this tumor.The mutation of genes BRCA-1 and BRCA-2 in 17q are identified in two separate types of hereditary carcinomas.Another hereditary type of OC is found in cancer family syndrome Lynch type 2 12 .As for environmental factors, it has been suspected that talc and asbestos may influence the onset of the disease, due to the fact that the highest incidence rates occur in highly industrialized countries.Some migrant studies have shown that when women from an undeveloped or developing country move to industrialized country develop OC 10 , while certain casecontrol studies pointed out that high intake of animal fat, alcohol and smoking may increase the risk 13 .
Regarding burden of OC in our country, official data revealed it as the 7th most frequent cause of cancer-related death as well as the 2nd most common cause in gynecological cancer deaths in the Serbian female population 7 .Furthermore, recent investigation showed statistically significant increase in OC mortality trend during the period 1976-2007 14 .Thus, the aim of this study was to determine the risk factors for OC in the population of Belgrade (Serbia).

Methods
This case-control study included 80 consecutive patients treated and followed in the Department of Gynecology and Obstetrics of the Clinical Center of Serbia and the Clinic of Gynecology and Obstetrics "Narodni Front" in Belgrade, between 2006 and 2008.Both of these hospitals are the national referral centers for OC in Serbia.All cases resided on the territory of Belgrade and had histologically verified diagnose of epithelial OC.The control group consisted of 160 women, double matched according to age (± 2 years) and municipality of residence.For the study period, the control subjects were recruited during regular gynecological checkups in the Public Health Center in corresponding municipalities.These women had no malignant tumors and/or hormone-dependent problems.All women, in both groups, signed the informed consent for the participation in the study.The research was approved by the Institutional Review Committee.Written informed consents were obtained from the study participants.
All the study participants were interviewed during their visits to the above mentioned institutions by two physicians using the same questionnaire to collect demographic information, as well as information regarding personal and family history, lifetime residence, particular lifestyle (smoking, alcohol and coffee intake), occupational exposures to radiation and chemicals, as well as reproductive history.Smokers were defined as persons who reported everyday smoking during a 60-day period prior to completing the questionnaire.To assess sport and recreation activities participants were asked if they do moderate activities for at least 10 minutes at a time, such as brisk walking, cycling, swimming, or any other activity that causes some increase in breathing or heart rate.In order to analyze the influence of specific exposure to the risk of the disease, we categorized variables according to the cutoff values.These values were determined based on the mean (± SD) level of variables investigated in the control group.All questions were referring to the 5-year period prior to the diagnosis or the corresponding period for the controls.This was supplemented and validated by an examination of the medical records.
Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated separately for each variable using univariate conditional logistic regression analysis.Variables that were related to OC at a significant level of p < 0.05, entered the final model of multivariate conditional logistic regression analysis to evaluate their independent contribution to the overall risk of OC.

Results
The average age of women in the study and the control groups was 56.1 ± 10.8 years and 56.7 ± 10.6 years, respectively.
Average income in patients was 245.52 ± 10.2 euros as opposed to the control group, where it was around 534 ± 146.5 euros.Lower family income ( 250 euros a month) was statistically significant risk factor for OC, with OR = 2.5 (95% CI 1.4-4.3),p = 0.001.Also, the living space in the study group it was in average 56.5 ± 23.0 m² while in the control group was 64.3 ± 67.3 m², thus the calculated OR was 2.0 (95% CI 1.1-3.6),p = 0.018, which was statistically significant.
The characteristics of OC cases and controls regarding education level and occupational history are presented in Table 1.There were no statistically significant differences in educational level, years of schooling, occupational and employment status between patients and the controls.
Analysis of diseases in the family history showed that only cardiovascular diseases were more frequently registered in the study group compared to the controls (OR = 2.6, 95% CI 1.5-4.1,p = 0.001).
Oral contraceptives use and other contraceptive methods (condoms, mechanical contraceptive devices) was highly statistically significantly more frequent among women in the control group (OR = 0.2, 95% CI 0.1-0.7,p = 0.005; OR = 0.1, 95% CI 0.01-0.5,p = 0.001, respectively) (Table 2).Hormone replacement therapy use was more frequent in the control group without statistical significance.Hormone therapy for any other reason including infertility treatment was more frequent among the OC patients but also without a statistical significance.
Women in the case group smoked for on average, 23.5 ± 8.9 years while in the control group 18.4 ± 6.8 years.The mean number of cigarettes smoked per day was 23.1 ± 6.4 in the OC group and 18.5 ± 6.4 among controls.Longer duration of smoking as well as higher amount of cigarettes (per day) were statistically significantly more frequent in the study group in comparison to the controls (Table 3).Coffee  intake was 5.6 times higher in women with OC (p = 0.023).Longer period of coffee consumption was also statistically more frequent in this group (OR = 3.8, p = 0.001).Tea consumption on daily basis had a highly statistically significant protective effect (OR = 0.3, p = 0.001).Body height and body mass index (BMI) showed that women in both groups did not differ in mass at the age of 18 (56.5 ± 5.3 kg vs 57.5 ± 7.7 kg, p = 0.340), nor within a 5-year prior to their illness (71.2 ± 7.4 kg vs 71.0 ± 11.5 kg, p = 0.888).However, the average body height at the age of 18 was statistically significantly higher among the OC patients (170.8 ± 4.7 cm) compared to controls (168.4 ± 6.7 cm) (p = 0.007).
Sport and recreation activities of the study participants are presented in Table 4.The patients with OC practiced sports for 6.3 ± 2.1 years, and the controls for 11.8 ± 9.9 years.According to the results of our study, sport and recreation activities were statistically significantly protective (OR = 0.2, p = 0.011; OR = 0.4, p = 0.019).The women in   the control group were practicing sports for more years and for more hours weekly than the cases, without a statistical significance.
Additionally, women in the study group were seated for 4.1 ± 3.1 h per day during their working time while for the women in the control group it was 4.6 ± 2.6 h (p = 0.366).The average time in sitting position in their leisure was 3.4 ± 1.4 h in the study group as opposed to the control group, where it was 4.0 ± 2.3 h (p = 0.033).A greater number of hours in sedentary position was statistically significantly more frequent in the study group than in the control one (OR = 1.8, 95% CI 1.0-3.1,p = 0.048).
All variables related to OC at a significant level of p < 0.05 using univariate logistic regression analysis were included in the model of multivariate logistic regression analysis.According to multivariate analysis the following factors were significantly negatively related to OC: oral contraceptive use (OR = 0.1, 95% CI 0.1-0.5, p = 0.009) and recreation activities (OR = 0.3, 95% CI 0.1-0.7,p = 0.007).

Discussion
In our case-control study conducted between 2006 and 2008 in Belgrade, Serbia, we included 80 OC patients and 160 controls.The results we found in this study indicate that the use of oral contraceptives, as well as sports and recreation activities statistically significantly decrease the risk of OC.
A strong evidence of negative association between oral contraceptives use and the occurrence of OC has been recorded in a number of studies [15][16][17][18][19] .Oral contraceptives have a long-term favorable effect on the OC risk 20 .Specifically, this phenomenon occurs due to the reduction in estrogen levels in the ovaries and prevention of the ovulation 15,21 .Even a short-term oral contraceptive use has been reported to reduce the risk 22 .Our results confirm this well-documented and defined association.
Even though the relationship between physical activity and the risk of hormone-dependent tumors (such as breast and endometrium cancer) have been known [23][24][25][26] , there is only a small number of studies regarding the effect of exercising on OC.The two studies in the US and China reported a decreasing risk of OC among women who exercise some kind of physical activity 27,28 .However, a Swedish prospective cohort study found no such evidence 29 .Being a hormone-dependent tumor, OC largely depends on oscilations of estrogen.Physical activity affects adipose tissue by reducing it, thus mobilizing hormone depots.Also, certain theories suggest that very strenuous physical activity induces late menarche, ammenorrhea and anovulatory cycles [30][31][32] .In addition, some authors 19,33 registered that older age at menarche has a protective effect upon OC, but our study did not show any link between these two variables.
In a pooled analysis of case-control studies Ness et al. 34 found no association between the use of fertility drugs and overall risk of OC which is in accordance with the results obtained in our study.On the other hand, systematic review and meta-analysis performed by Greise et al. 35 showed that both menopausal estrogene and progestin therapies are the risk factors for OC.
Many authors [15][16][17][18][19] have reported nulliparity to be strongly associated with the ccurrence of OC, but in our sample of patients no significant difference in parity between the cases and the controls was found.Our patients also brestfed for shorter period.Moorman et al. 15 reported breastfeeding to be protective factor for the occurence of OC.They had lower family income and were of less educational level which is in concordance with the findings of El-Khwsky et al. 16 in Egypt and Song et al. 18 in China, whereas Zhang et al. 36 from the USA reported opposite results.
The women with OC were taller at the age of 18, but as for BMI no relationship was established.Moorman et al. 15 registered body height and BMI of 35 and over to be the risk factors for this tumor.Greer et al. 22 reported that women who had greater both recent weight and weight at the age of 18 were at higher risk of OC.Beehler et al. 37 showed that obese, premenopausal women have 2 times more chance of developing OC.However, they have not recorded any association between BMI in postmenopausal women and OC, which corresponds to our findings, that BMI does not carry any risk of this disease.
Smoking is often associated with many types of cancers 33 and our research confirmed that the cases smoke more than their controls.However, controls drank more tea and coffee on daily basis.Data on this issue in literature is controversial.Some studies 18,36 have shown protective effect of the green tea, as opposed to black tea.Our study did not divide teas into categories.And that neither caffeinated or decaiffenated coffees were associated with the risk of OC, which opposes to our findings 18 .
One can criticize that the recall bias may have influenced the results in this case-control study and the size of our study groups.In addition, population controls may have been more appropriate and therefore our study is certainly subject to selection bias.In the present research we did not look for specific dietary patterns of our patients, even though certain articles suggest micronutrients such as calcium, vitamin E and beta-carotene have protective effect upon occurrence of OC 38,39 and that meat and fat are associated with an increased risk of OC 40 .

Conclusion
Based on data obtained in our study, oral contraceptives use and physical activity are independent protective factors for OC.Overall, this is the first epidemiological study on risk factors for this cancer in our country.In future, attention should be paid on finding a larger sample, which would select participants throughout the country, not only from one major center (Belgrade) and broaden the scope of the specific questionnaire.

Table 1 Characteristics of the ovarian cancer cases and the controls regarding education level and occupational history
OR -odds ratios; CI -confidence intervals.

Table 3 Smoking and coffee consumption
*Current and former smokers; OR -odds ratios; CI -confidence intervals.