Introduction: Obesity and Cancer Epidemiology

Geographic and socioeconomic differences in cancer trends and incidences indicate that environmental factors affect the development of cancer.


Continue Reading

Changes in cancer epidemiology over the last number of decades cannot be accounted for solely by changing population demographics – including increased life expectancy, increased population numbers, or improved screening, detection, and treatment leading to increased numbers of people living with a cancer diagnosis.

Since not enough time has elapsed in order for significant genetic changes to account for the increased cancer incidence, it is thought that it is attributable to environmental factors. There has been an evolution in society’s structure from hunter–gatherer to peasant–farmer and increasingly urbanized to industrialized.

Doubtless there are have been a number of closely interrelated changes to body composition, physical activity, energy balance, and diet, and it is these environmental changes which are thought to be the cause of the alteration in cancer epidemiology.

The most compelling evidence for the association between environmental factors and cancer incidence comes from migrant studies.

Migration of populations, from rural to urban environments and within and between countries has been used to show how migrant populations display changes in cancer incidence similar to their adopted country within one to two generations.1

For example, a migrant Japanese population which moved to Hawaii developed a threefold increase in breast cancer incidence in the first generation of migrants and subsequently developed up to a fivefold increase in the second generation.2

Similarly, the colorectal cancerincidence increased by fourfold in the first generation, butdid not increase with subsequent generations, and the gastriccancer incidence, which is high in Japanese populationswithin Japan, decreased by almost 50% in the first generationand fell again in the second generation.

Thus, theseJapanese migrants developed cancer in the same pattern astheir Caucasian Hawaiian neighbors than those of the sameethnicity in Japan. Similar changes among other migrantpopulations have been replicated in Australia,3 Canada,4and the United States.5

Cancer incidence rates generallybecome similar to those of their adopted country in secondgeneration immigrants – long before any new geneticdifferences could arise.Interestingly, the increase in cancer due to environmentalfactors is similar in populations with clearly identifiableheritable risk factors for cancer development than those withsporadic cancers.

For example, in an Icelandic population therisk of breast cancer by age 70 in BRCA-2 carriers increasedfrom 18.6% in 1920 to 71.9% in 2002. This was matched bya parallel increase in the sporadic breast cancer rate, whichrose from 1.8% to 7.5%.6Excess nutrition and low physical activity can manifestitself as obesity.

The prevalence of obesity has increasedrapidly over the past 30 years.7 Being overweight or obese isthe most prevalent body composition in some countries, forexample, in the United States, where 72% of men and 67%of females are overweight or obese.8

Rates are increasing in asimilar fashion in Western Europe, with 65% of men and 56%of women being overweight or obese in the United Kingdom.9This trend shows no signs of abating as obesity rates areincreasing among children,10 and overweight children tendto become overweight adults.11

In countries where sedentary lifestyles and high-energyfoods are abundant, it is easy to see how energy intakeexceeds that expended. It has been estimated that ingestion of5% more calories than expended may result in an accumulationof 5 kg of adipose tissue in a single year.12

Humans arepoorly able to distinguish foods with a high-energy contentper mass, such as those high in fat and sugar, and this oftenleads to passive increases in calorie consumption, whichis deposited as excess adipose tissue.13

In a recent analysisof the European Prospective Investigation into Cancer andNutrition (EPIC) study, avoiding inactivity led to an equalreduction in all-cause mortality risk as did avoiding a highwaist circumference, indicating that physical inactivity playsan independent role in mortality risk even in overweightpatients.14 

Morbidity and mortality, attributable to obesity, risesharply with body mass index (BMI) .30 kg/m2,15 with therisk of premature death doubling as BMI exceeds 35 kg/m2.16BMI alone is a poor measure of obesity as it may poorlyreflect body composition, with raised BMIs found in patientswith high muscle bulk.17

It also does not reflect differencesin the site of fat distribution, and it is thought that visceralor central adiposity is more often associated with metabolicdysregulation in obesity.18,19 For a given BMI, men are atgreater risk of mortality than women,20 possibly relatingto the increase in visceral adiposity in men versus women,which is not reflected by BMI.

Conversely, energy restrictionin laboratory animals has been shown to decrease spontaneoustumor occurrence in a meta-analysis.21 Mice geneticallysusceptible to the development of colon cancers have areduced incidence of polyps when fed a calorie-restricteddiet.22 Patients who had been hospitalized for anorexianervosa prior to the age of 40 have been shown to have a23%–76% decreased incidence of breast cancer (dependingon parity).23

Thus, the optimal dietary approach to cancerrisk reduction is thought to be CRON – calorie restrictionwith optimal nutrition.24Epidemiological studies have provided convincing evidencefor the association of obesity with cancer.25–27

TheWorld Cancer Research Fund used a standardized approachin analyzing the evidence and concluded that there is convincingevidence of association between obesity and esophagealadenocarcinoma and pancreas, colorectum, breast (postmenopausal),endometrium, and kidney cancer.28

The largest meta-analysis to date includes 282,000 patientsfrom prospective observational studies with over 133 millionperson-years of follow-up.29 This comprehensive analysisshows that high BMI is associated with an increasedincidence of many types of cancer.

The association is modestwith risk estimates of 1.1–1.6 per 5 kg/m2 incrementalincrease in BMI. This 5 kg/m2 increase in BMI correspondsto a 15 kg weight gain in men and 13 kg in women, with anaverage BMI of 23 kg/m2.

Longitudinal epidemiological studies implicate obesityas a causal factor in the development of cancer. Further epidemiologicalevidence of causality comes from prospectivestudies of patients undergoing bariatric surgery with sustainedlong-term weight loss, such as those included in the SwedishObese Subjects (SOS) study.

At 10 years follow-up, the SOSstudy reports a reduced risk of developing cancer, whichseems to be present in women alone (women: relative risk0.58, 95% confidence interval [CI] 0.44–0.77 versus men:relative risk 0.97, 95% CI 0.62–152).

Whether sex-specific or due to insufficient length of follow-up and/or sample size, it remains an area for future study.30