Risk Factors for Breast Cancer in India: an INDOX Case-Control Study

1.  Study Investigators

Principal Investigators: Oxford

  • Dr Toral Gathani
  • Dr Raghib Ali
  • Professor Dame Valerie Beral

Principal Investigator: India

  • Professor Vinod Raina, AIIMS, New Delhi

Local Principal Investigators: India

  • Prof. T.S. Ganesan, AIMS, Cochin
  • Dr Sara Beela Mathew, RCC, Trivandarum
  • Dr Rajkumar, GKNMH, Coimbatore
  • Dr Chacko, CMC, Vellore
  • Dr Raghunadharao, NIMS, Hyderabad
  • Dr Loknatha, KMIO, Bangalore
  • Dr Shilin Shukla, GCRI, Ahmedebad
  • Dr Hemant Malhotra, BCC, Jaipur
  • Dr Rajiv Prasad, IGIMS, Patna
  • Dr Ravi Kannan, CCC, Silchar


2.  Summary

Breast cancer is the most common cancer diagnosed in women worldwide with over 1.3 million new cases per year.  There is a wide variation in the geographical burden of the disease with the highest incidences seen in the more developed regions of the world and the lowest incidences observed in the least developed regions.  More recently the incidence of breast cancer has been observed to be increasing in low income countries and data suggests that over the next twenty years the majority of the increase in the worldwide burden of the disease will be due to rising incidences in these countries.

India is undergoing a period of dramatic social and economic change.  Cancer is now the second leading cause of death in Indians after cardiovascular disease.  Amongst women cervical cancer is still the most frequently diagnosed cancer but breast cancer is now the most commonly diagnosed cancer in urban Indian women.  The reasons for the recent observed increase in incidence of breast cancer in the Indian population are not clearly understood but thought to be largely explained by ‘westernisation’ of lifestyles and changes in reproductive behaviour.

By conducting a large scale multicentric case control study within India we hope to understand further the risks for breast cancer in this population which can then inform public health strategies to try and reduce the burden of disease.  The Indian population also provides  a unique opportunity to investigate the role of risk factors such as vegetarianism which are highly prevalent in this population group compared to the West.

3.  Background

Breast cancer is the most common female cancer worldwide with an estimated 1.38 million new cancer cases diagnosed in 2008 representing 23% of all cancers in women.  It is estimated that by 2030 the global burden of breast cancer will increase to over 2 million new cases per year. Furthermore it is estimated that this increase in cases will be largely due to increasing incidence in developing regions of the world.

Figure 1: IARC estimation of global burden of breast cancer in 2030


The world age standardised incidence rate for breast cancer is 39 per 100000. The incidence rates show marked geographical variation from 27.3 per 100000 in less developed countries to 66.4 per 100000 in more developed countries.  However the population ratio of 1:4 in developed to developing countries means that breast cancer is the most common female cancer in both the developed and developing regions with approximately 690000 new cases estimated in each region in 2008.  The highest incidence rates are observed in Western Europe (89.9 per 100000) and the lowest in Eastern Africa (19.3 per 100000).

Figure 2: Estimated incidence of breast cancer worldwide in 2008

map showing estimated incidence of breast cancer worldwide in 2008

Breast cancer incidence rates have decreased in the USA and parts of Western Europe since the early 2000s.  The drop in rates has largely been attributed to decreasing use of hormone replacement therapy in these populations.  In contrast to the incidence trends in the developed world, breast cancer incidence and mortality rates have been increasing rapidly in less developed regions.

Breast cancer is the most frequent cause of cancer related death for women in both developed and developing countries.  The estimated number of deaths globally due to breast cancer in 2008 was 458367.  Mortality rates show less geographical variation compared to incidence rates because of the more favourable survival of breast cancer in high-incidence developed regions. The world standardised mortality rate for breast cancer is 12.5 per 100000 and ranges from 17.5 per 100000 in Western Europe to 6.3 per 100000 in Eastern Asia.

Figure 3: Age standardised incidence and mortality rates for breast cancer worldwide per 100000 in 2008

3.1.  Breast Cancer Incidence and Mortality in India

The incidence of breast cancer is rising in India and is now the second most commonly cancer diagnosed in women after cervical cancer.  It is estimated that in 2008 there were 115251 new cases of breast cancer with an age standardised incidence rate of 22.9 per 100000.  It is estimated that by 2030 the number of new cases of breast cancer in India will reach just under 200000 per year (Ferlay J, 2010).

Figure 4: Estimation of number of incident breast cancer cases in India diagnosed in 2030

Chart showing estimation of number of incident breast cancer cases in India diagnosed in 2030

Data from National Cancer Registry Programme shows that in all urban areas of India breast cancer has now surpassed cervical cancer as the most frequently diagnosed cancer in women.  The most recent data available from the National Cancer Registry Programme show a wide variation in age standardised incidence rates observed between rural and urban populations ranging from 36.1 in Bangalore to 7.2 in Sikkim state.

Figure 5: Age standardised incidence rates for breast cancer in women NCRP 2006-8

Chart showing age standardised rates for breast cancer in women across India NCRP 2006-8

The age standardised mortality rate for breast cancer in India is 11.1 per 100000 (12.5 per 100000 globally).  In common with other developing regions mortality rates for breast cancer in India are high in comparison to incidence rates.  Poor survival may be largely explained by lack of or limited access to early detection services and treatment.

3.2.  The Risk Factors for Breast Cancer

The risk factors for the development of breast cancer in Western populations are well established but data regarding the effect of these known risk factors within developing world populations are less clear. Breast cancer risk factors can be considered in terms of those which relate to non-reproductive factors and those that relate to reproductive factors. The higher incidence of breast cancer observed in the developed world as compared to the developing world reflect the long-standing high prevalence of reproductive factors associated with an increased risk of breast cancer including early menarche, late child bearing, fewer pregnancies, reduced duration of breast feeding, use of hormone replacement therapy as well as increased detection through mammographic screening.  The reasons for the low incidence of breast cancer among Indian women and the increasing incidence observed in recent years are not completely understood although likely to be explained by reproductive and lifestyle factors.

3.2.1.  Non-reproductive Factors

The incidence of breast cancer increases with increasing age with 80% of all breast cancer occurring in women over the age of 50 in the UK. The average age of presentation for breast cancer in the Indian population is widely reported to be around 10 years younger compared to the developed world and is the result of the age structure of the Indian population which is a bottom-heavy (predominantly young) pyramid.

The role of diet in the aetiology of breast cancer remains controversial.  Interest in the role of diet in the aetiology of breast cancer is stimulated by the observation of the lower incidence of breast cancer in Asian populations where intake of animal products is lower than that of western populations.  Worldwide, a large number of individual follow vegetarian diets but in most countries vegetarians comprise only a small proportion of the population. India is a notable exception because a substantial proportion of the population (perhaps as high as 35%) follow a vegetarian diet and have done so for many generations.  Studies to date exploring the relationship of diet and breast cancer incidence and mortality have not demonstrated any difference between vegetarians compared to non-vegetarians. A study specifically examining the effect of vegetarian diet on subsequent risk of breast cancer in British Asian women showed that lifelong vegetarians had a slight reduction in the risk breast cancer development compared to meat eaters but the result did not achieve statistical significance (RR 0.77 (95% CI 0.5-1.2).  Specific studies in India examining this relationship are few and of small size and show inconsistent findings.  An older study of 689 breast cancer patients in Mumbai found no association between vegetarian diet and breast cancer risk  whereas a study in Kerala of 264 women with breast cancer showed an increased risk of breast cancer associated with a non-vegetarian diet (OR 1.82 (95% CI 1.11-2.98)). A large scale case-control study such as this one to investigate the relationship of breast cancer incidence and diet in the Indian population would be of interest.

The two diet related factors which have been shown to have an effect on breast cancer incidence within western populations are obesity and alcohol intake.  The incidence of breast cancer in obese postmenopausal women is twice that of the non obese and is related to increased levels of circulating oestrogens due to the peripheral aromatisation of oestrogen in adipose tissue whereas obese pre-menopausal women have a 20% reduction in cancer risk.  Increasing body mass index (BMI) is increasingly becoming a concern in many low-resource countries, particularly in urban India with studies reporting an increase in BMI in urban women compared to women living in rural areas.  A large case control study conducted in South India has shown an increased risk of breast cancer in both overweight pre-menopausal (OR 1.33 (95% CI 1.05-1.69)) and post-menopausal women (OR 1.29 (95% CI 1.00-11.66)).  The findings in pre-menopausal women of an increased risk of breast cancer associated with increased BMI is not consistent with western data.

The Collaborative Group on Hormonal Factors in Breast Cancer in a large collaborative reanalysis of the worldwide data have shown that the relative risk of breast cancer is increased (RR 1.32 (95% CI 1.19-1.45)) in women who reported drinking alcohol compared to women who reported no alcohol intake for an intake of 35-44 g per day alcohol.  Alcohol intake amongst Indian women is traditionally low but with changing lifestyles within the urban population it is possible that alcohol consumption has increased and this may in part contribute to increased breast cancer rates but no recent data are available.

3.2.2.  Reproductive Factors

There is a large body of evidence from both experimental and epidemiological studies which points to a major influence of ovarian hormones on breast cancer risk with the lifetime risk of developing breast cancer is related to overall oestrogen exposure.

Age at menarche is an important determinant of subsequent breast cancer risk.  Estimates from a pooled analysis of the results of 21 studies shows that for each additional year age at menarche is postponed, premenopausal and postmenopausal breast cancer risk decreases by 9% and 4% respectively.  The median age of menarche worldwide is 14 years with a range from 11-18 years and reported average later age of onset in Asian populations compared to the West.  A study in South Indian women showed that the risk of both pre-menopausal and postmenopausal breast cancer decreased with delay of the onset of menarche but not significantly so. Two smaller studies, one of Mumbai women and one of Delhi women also showed no significant association between age of onset of menarche and subsequent breast cancer risk.

Late menopause increases the risk of breast cancer. Women who have undergone the menopause have a lower risk of breast cancer than pre-menopausal women of the same age and childbearing pattern.   Risk increases by almost 3% for each year older at menopause (natural or induced by surgery), so that a women who has the menopause at 55 rather than 45, has approximately 30% higher risk.  The South Indian study showed an increased risk of breast cancer in woman who became menopausal after the age of 50 (RR 1.87 (95% CI 1.26-2.78)) but there is little other published data.

Child bearing and breast feeding are of importance when considering subsequent breast cancer risk. The younger the woman is when she begins childbearing, the lower her risk of breast cancer. The relative risk of developing breast cancer is estimated to increase by 3% for each year of delay. Childbearing reduces the risk of breast cancer and the higher the number of full-term pregnancies, the greater the protection. The risk of breast cancer reduces by 7% with each full-term pregnancy, and overall women who have had children have a 30% lower risk than nulliparous women

Women who breastfeed reduce their risk compared with women who do not breastfeed. The longer a woman breastfeeds, the greater the protection: risk is reduced by 4% for every 12 months of breastfeeding.  There is also evidence that the reduction in risk of breast cancer with childbirth, and higher risk with later age at first full-time birth, may be limited to oestrogen-receptor-positive tumours.  

Data from Indian studies are less clear. The largest study in India to examine the relationship between breastfeeding and breast cancer risk has shown mixed results.  Increasing duration of breast feeding was associated with a significantly decreased risk of pre-menopausal breast cancer but no effect was seen in women with postmenopausal breast cancer.  With regard to parity, the researchers observed that as the number of children a woman has increases, the subsequent risk of breast cancer decreases.  Conversely, the Delhi study showed a protective effect of breast feeding of greater than 6 years duration on breast cancer risk but no significant protective effect related to parity and the Mumbai study showed a non-significant protective effect for breast feeding.  These studies are of small sample size and thus probably underpowered to detect significant differences.

The data from India overall suggests that the role of known risk factors on the development of breast cancer within the Indian population is unclear and a large multicentre study would be of benefit to try and understand the shifting trends in breast cancer incidence in this population.

4.  Aim

The primary aim of this case control study is to explore the association between reproductive patterns (e.g. breastfeeding and parity) and dietary patterns highly prevalent among Indians (e.g. lifelong vegetarianism) and the subsequent risk of breast cancer

5.  Methods

  • The study design is an interviewer administered questionnaire based case-control study taking place in eleven of the INDOX collaborating centres in India.
  • Data analysis will take place in both Oxford and India.
  • Anticipated duration of study is 24 months from the commencement of patient recruitment including the pilot phase of the study.

This study will begin recruiting patients in India in Spring 2011.