CHAPTER – 1

1.1 Introduction

Colorectal cancer is considered to be a major cause of cancer morbidity and mortality. It accounts for over 9% of all cancer incidences; this makes it the third most common cancer cause with nearly 1.4 million new cases diagnosed in a year (

WCRF, 2012

) and the fourth most common cause of death (

WHO, 2002

). Both men and women are affected almost equally for colorectal cancer; 9.4% in men and 10.1% in women from all incident of cancer. However, it has a clear geographical variation which is not uniformly distributed throughout the world. It has been estimated that, the developed countries with a western culture accounts for over 63% of all colorectal cancer cases (

Fatima AH, 2009

). Though data available to show the clear situation in developing countries, growing evidences show that the problem would be also a concern for the low and middle income countries. The International Agency for Research on Cancer (IARC) was estimate that in 2008 colorectal cancer (CRC) is the 5th most common cancer in SSA (

Ferlay J, 2010 and


Alice G et.al, 2012

)

Several risk factors can be linked with the incidence of colorectal cancer. Age and hereditary factors are the most important factors on which an individual’s cannot able to manage. The probability of being affected by colorectal cancer is increases after the age of 40. More than 90% of colorectal cancer cases reported among people in the age greater than 50 and older

(Fairley TL , 2006)


.

In addition, a large number of environmental and behavioral risk factors can also contribute for the development of colorectal cancer; among these dietary factors are the major one

(


Fatima AH, 2009


).

With the help of different dietary assessment methods such as food recalls and food record, wide range of epidemiological studies revealed that diet can strongly influences the risk of colorectal cancer, and changes in food habits might reduce up to 70% of this cancer burden. Diets with high in fat and calories, especially animal fat, protein rich, low level of calcium and vitamin D. are some of the nutrients which can be a risk factor for colorectal cancer. High meat consumption, mainly red meat and processed meat, is mainly associated with the development of colorectal cancer (

Fatima AH and Robin P, 2009;


Sandhu MS et.al, 2001; Norat T et.al, 2002; Larsson SC, 2006, Raphaëlle L, 2008)

. On the other hand, those people who eat diets low in fruits and vegetables may have also a higher risk of developing colorectal cancer

(


Paul T et.al, 2001


).

For free living person, accurate estimation of routine dietary intake is very challenging in the study of diet and disease relationships (

Jackson et. al, 2011

). Moreover, there is no single method of assessment which is valid and optimal under all conditions to assess individual’s habitual dietary intake, but because of its ability to rank subjects according to their intake and also it is relatively inexpensive, food frequency questionnaires (FFQs) are often used in most epidemiological studies to assess usual dietary intake and disease development. In addition, Food frequency questionnaires (FFQs) have been used to assess long-term dietary intakes and an important exposure factor for the disease conditions (

Sofi F et.al, 2008, and Streppel M, et. al, 2013

). Therefore, the use of food frequency questionnaires has advantage over the other methods such as 24-hour recalls and food records since these methods do not reflect past diet intake or usual intake and they are also expensive (

Jackson et. al, 2011

).

Even though the Food frequency questionnaires (FFQs) is a more practical tool for assessing individual diets in large cohorts, it has also more associated with measurement errors than 24-hour recalls and dietary record assessment methods. In consequence, this can bias the relative risk estimates of diet and disease risk (

Lin L et.al, 2013

). One approach to improve effect estimates is through validation studies, in which individual diets by questionnaire is compared with a more precise and accurate method such as weighted dietary records (

Jaceldo-Siegl et. al, 2009

).

The influence of dietary habit is central for the development of colorectal cancer but little is known about validity of dietary exposure assessment tools. The purpose of the study is, therefore, to evaluate the validity of the FFQ that will be used in assessing the relationship between dietary intake and

colorectal cancer

.

1.2 Objective of the study

• To evaluate the food frequency questionnaires used to assessing the risks of dietary intake (at food group and nutrient level) for colorectal cancer with the three days weighted diet record method

CHAPTER – 2

2.1 Literature Reviews

2.1.1 The need of validation studies for food frequency questionnaires

The role of dietary factors in the etiology of several cancers has been extensively investigated over the last few years including on colorectal cancer (

Bazensky I, Shoobridge-Moran C, Yoder LH, 2007

). However, accurate estimates of habitual dietary intake remain a challenge in the study of diet-disease relationships (

Jackson et. Al, 2011

). This is because dietary assessments could be affected by a number of factors such as motivation to complete assessments and reporting bias related with unstructured eating patterns(

Livingstone MB et. al, 2009

). Besides these, design of the study, outcomes of interest, and resources availability need to be considered when selecting an appropriate dietary assessment tool for a particular study(

Jyh Eiin Wong et. al, 2012

).

In epidemiological studies, Food-Frequency Questionnaire (FFQ) is often chosen assessment method for estimating dietary intake and can provide valid and reliable estimates of usual dietary intake in a variety of populations (

Navarro A, et. al,


2001

) where as the other possible and precise methods such as food records and 24-hour recalls do not reflect past diet or usual intake and are generally expensive (

Jackson et. al, 2011, Lin L et.al, 2013

).

The FFQ is easy to administer, has relatively low cost, and provides a rapid estimate of usual food intake. Due to having this advantages it is more practical and frequently used dietary assessment tool for assessing individual diets in large cohorts. However, compared to other dietary assessment methods it has more associated with measurement errors that usually overestimate relative risks in studies of diet and disease risk (

Jaceldo-Siegl et. al, 2009; Zulkifli SN, 1992; Fraser GE, 2003; Lin L et. al, 2013; Streppel M et. al, 2013)

.

Therefore, nutritional values reported from FFQ data are subjected to both systematic and random errors that can significantly affect the design, analysis, and interpretation of nutritional epidemiologic studies (

Carroll RJ, et. al, 1997

). In addition, due to possible differences in design characteristics of the FFQ, e.g. the number of food items included, mode of administration, and also the need of cognitive process for portion size estimation, it’s validity can be affected (

Jyh Eiin Wong et. al, 2012)

. Furthermore, the validity of the same FFQ may vary from one population to the other (

Streppel M. al et, 2013

).

The above rationale can lead to the agreement that evaluation and validating of a FFQ is important in studying diet – disease associations. Thus validation studies need to be performed to evaluate the level to which the FFQ agrees with the subjects’ true dietary intake (

Cade JE et. al, 2004

). Moreover, validation studies can be also carried out to assess the level of measurement error associated with the FFQ (

Streppel M et al, 2013 and Cade JE et. al, 2004

).

Providing the fact that there is no single method which is completely free from possible biases in dietary assessment, one approach to validate the estimates of effects through the help FFQ is by comparing with a more precise method (reference method), such as weighted food record, multiple dietary recalls and use of biomarkers (

Jaceldo-Siegl et. al, 2009 and Lin L et.al, 2013

). However, memory (recalls) and nutrient data are still influence estimates from the reference method (

Jackson et. al, 2011, and Daures JP et. al, 2000

).

Generally, due to differences in food supply and dietary habits, there is no universally accepted FFQ that can be used for all populations in all situations. Elucidation of diet–disease relationships requires dietary assessment methods which can sufficiently describe and quantify intakes, minimize errors and provide precise estimates of variability between individuals or groups(

Kaaks R et. al, 1997; Carroll RJ e t.al 1997 and Lin L et.al, 2013

). Therefore, it is vital to make sure that any FFQ must be reliable and valid to be used in the population of interest, need to be designed to meet the aims of study populations and has contain an up-to-date list of foods(Jyh Eiin Wong,et. al, 2012).

2.1.2 Incidence and mortality of colorectal cancer

Colorectal cancer is one of the major cause of morbidity and mortality. Globally it accounts for over 9% of all cancer incidences; the third most common cancer and the fourth most common cause of death. Men and women are almost equally affected by colorectal cancer; it represents 9.4% in men and 10.1% in women (

Jodi D Stoocky, et al, 1996


).

However, is not uniformly distributed throughout the world; it has a clear geographical variation. Mainly, colorectal cancer is a disease of developed countries with a western culture. It is estimated this region accounts for over 63% of all cases. Worldwide mortality attributable to colorectal cancer is approximately half that of the incidence. In 2008, over 1.2 million new cases and 608,700 deaths estimated to have occurred (

Fatima AH, 2009 and Dagfinn A, 2011

).

The WHO report in 2012 also showed that, in 2008 among the deaths of an estimated 7.6 million people of all cancer cases colorectal cancer (CRC) accounted for over 600 000 of those deaths, with 70% occurring in low– and middle–income countries (WHO, 2012 and Meetoo D, 2008). This indicates that though the number of cases of CRC in SSA is thought to be very low in comparison to those diagnosed in the Western world, it constitutes a significant proportion of the cancers in this region (Alice G et.al, 2012).

2.1.3 Risk factors for colorectal cancer

Colorectal cancer is widely considered to be an environmental disease; include a wide range of cultural, social, and lifestyle factors which are associated with the incidence of colorectal cancer. Age and hereditary factors are those that an individual cannot control and other modifiable environmental and lifestyle risk factors are also plays an important role in the development of colorectal cancer (Fatima A. Haggar, Robin P. Boushey 2009).

Among the modifiable risk factors, diet habit is a major one and strongly influences the risk of colorectal cancer, and changes in dietary habits might also decrease up to 70 percent of this cancer burden. Diets with high fat content, especially animal fat, considered to be a major risk factor to cause colorectal cancer. However results of meta analysis supports the hypothesis that only intakes of red and processed meat are convincing dietary risk factors for colorectal cancer (Dagfinn Aune, 2011). In addition, different studies also revealed that the association of CRC risk with

processed red meat

may be stronger than that of

fresh red meat

(Sandhu MS et.al, 2001; Norat T et.al, 2002; Larsson SC, 2006 and Raphaëlle L, 2008).

The effect of fat consumption as possible etiologic factor for colon cancer is linked to favoring the development of a bacterial flora which capable of degrading bile salts to potentially carcinogenic compounds. The presence of heme iron in red meat believed to be the main mechanisms for the positive association in the development of colorectal cancer. In addition, exposing it to high temperatures could resulting the production of compounds such as heterocyclic amines and polycyclic aromatic hydrocarbons (Fatima A. Haggar, Robin P. Boushey 2009 and Genkinger JM, Koushik A, 2007).

MS Sandhu et al in 2001 found that red and processed meat could contribute for colorectal cancer; a daily increase of 100 g of red meat and 25g of processed meat was associated with a 14% (OR;1.14, 95% CI:1.04 to 1.25) and 49% (OR; 1.49, 95% CI: 1.22 to 1.81) increased risk of colorectal cancer respectively (Sandhu MS et.al, 2001). On the other more recent follow up study by Cross, A.J et.al, the hazard ratios (HR) and 95% confidence intervals (95% CI) comparing the fifth to the first quintile for both red (HR; 1.24; 95% CI, 1.09-1.42; P < 0.001) and processed meat (HR; 1.16; 95% CI, 1.01-1.32; P = 0.017) intakes indicated an elevated risk for colorectal cancer (Cross, A.J et.al, 2010).

Norat, T in 2001 also tried to conduct meta-analysis of articles published during 1973-99 in order to look the risk of consumption of red and processed meat for colorectal cancer. High intake of red and processed meat was associated with significantly increase risk of colorectal cancer. Average Relative Risk and 95% confidence intervals (CI) for the highest quantile of consumption was 1.35 (CI: 1.21-1.51) for red meat and 1.31 (CI: 1.13-1.51) for processed meat (Norat, T., et.al, 2002). A similar effect of red and processed meat also found in other meta-analysis of prospective studies published through March 2006 by Larsson SC and Wolk A in 2006. This meta-analysis of prospective studies support the hypothesis that high consumption of red meat and of processed meat is associated with an increased risk of colorectal cancer (Larsson, S.C. and Wolk, A., 2006).

In addition, some studies suggest that people who eat a diet low in fruits and vegetables may have a higher risk of colorectal cancer. Differences in dietary fiber intake might have been also responsible for the observed geographic differences in the incidence rate of colorectal cancer (Janout V, and Kolla´rova´ H. 2001 and Fatima AH, 2009).

Age is an important non modifiable factors in the development of colorectal cancer. As the increases the likelihood of colorectal cancer diagnosis increases as well. Mostly its diagnosis is common after the age of 40, then rising sharply after age 50. More than 90% of colorectal cancer cases occur in people aged 50 or older. The incidence rate is more than 50 times higher in persons aged 60 to 79 years than in those younger than 40 years. However, colorectal cancer appears to be increasing among younger persons (Ries LAG, et al.. 2008, and Fatima AH, 2009).

2.1.4 Dietary protective factors for colorectal cancer; Vegetables, Dietary fiber, whole grains

Given the roles of the colon and rectum as conduits for ingested food and the many potentially anti carcinogenic substances contained in fruit(

8

),vegetables(

8

),and cereals(

9

),these food groups are among the most widely studied in relation to colorectal cancer risk (Steinmetz KA, Potter JD. 1991, and Slavin JL, Martini MC et.al, 1999).

Although the majority of studies have shown an inverse association between fruit and vegetable consumption and colorectal cancer risk, some prospective cohort studieshave also obtained inconsistent results; no association between fruit and vegetable consumption and the risk of having or developing a colorectal neoplasm (Paul T et.al, 2001).

Results from a meta-analysis of 13 case–control studieshave suggested that increased dietary fiber intake is associated with decreased risk of colorectal cancer, although some other prospective cohort studies do not supporting such an association (Steinmetz KA et.al, 1994). Nonetheless, evidence from animal studies and few clinical trials continues to suggest that cerealsespecially wheat bran, contain substances, such as fiber, phytic acid, various phenolic compounds, lignins, and flavonoids, that might lower the risk for colorectal cancers.Cereal fiber might also bind carcinogens and modify glycemic index(Slavin JL , 1999, and Paul T et.al, 2001)

In a Paul T et.al study in 2001, they were tried to associate the role of fruit and vegetable consumption and colorectal cancer. In this study 61, 463 women were followed for an average of 9.6 years and 460 incident cases of colorectal cancer were observed (291, 159 and 10 cases of colon, rectal and both sites respectively). In the entire population of this study, total fruit and vegetable consumption was inversely associated with colorectal cancer risk. The association was also showed that there is a dose–response effect; more evident among individuals who consumed the lowest amounts of fruit and vegetables was observed. In particularly, those individuals who consumed less than 1.5 servings of fruit and vegetables per day had a 1.65 relative risk (95% CI = 1.23 to 2.20;P = .001) for developing colorectal cancer compared with individuals who consumed more than 2.5 servings. However, no association between colorectal cancer risk and consumption of cereal fiber was observed in this study. In conclusion this particular study revealed that individuals who consume very low amounts of fruit and vegetables have the greatest risk of colorectal cancer (Paul T et.al, 2001)