Disparities in Cancer Risk Factors and Cancer Screening in the United States Part 1

Much progress has been made over the past three decades in understanding, preventing, detecting, diagnosing, and treating cancer (Curry, Byers, & Hewitt, 2003; Jemal, Ward, & Thun, 2010). Cancer prevention, as a public health strategy based on research evidence, is the first line of defense in reducing the number of deaths resulting from cancer. It includes medical approaches (i.e., use of recommended cancer screenings) as well as environmental and behavioral interventions to modify risk factors. While many factors contribute to cancer risk, the chief behavioral (modifiable) risk factors are tobacco use, obesity, physical inactivity, and diet (Curry et al., 2003; World Cancer Research Fund/American Institute for Cancer Research, 2007). These behavioral risk factors also have a large impact on the incidence of other major chronic illnesses, such as cardiovascular disease and diabetes (Curry et al., 2003; World Cancer Research Fund/American Institute for Cancer Research, 2007). Moreover, comprehensive reviews have documented disparities in these risk factors among racial, ethnic, and socioeconomically vulnerable populations (Ward et al., 2004), and by geographic region ("Vital Signs," 2010). Though not fully understood, the causes of disparities in behavioral risk factors for cancer appear to be multifaceted and complex; experts suggest that cancer disparities are partly due to lack of access to medical care, including preventive care and state-of-the-art cancer services, environmental factors that deter adoption of healthful behaviors, or that facilitate unhealthful ones (e.g., the higher prevalence of fast-food outlets in segregated Black neighborhoods), and the interplay of low socioeconomic class, culture, and social injustice. On the other hand, inherent biological characteristics are not considered key factors in influencing cancer disparities (Ward et al., 2004). In this topic, we discuss the major (modifiable) risk factors for the most common cancers in the United States as well as disparities in their prevalence among sociodemographic groups (i.e., race/ethnicity, gender, and socioeconomic status). We also summarize research on interventions with the potential to reduce disparities among special populations (as defined by race/ethnicity, sexual orientation, etc.) in tobacco smoking, the single preventable risk factor responsible for the majority of cancer cases.

Various national survey data are routinely used to monitor health status, medical conditions, and health services utilization in the United States. These provide prevalence estimates (proportion of the population) of behavioral risk factors for groups belonging to the major sociodemographic categories (age, race/ethnicity, gender, income, and education). Three national government-sponsored surveys are the National Health Interview Survey, the National Health and Nutrition Examination Survey, and the Behavioral Risk Factor Surveillance System. It is important to recognize some limitations of estimates from national surveys, however, that may result in less accurate information for some racial/ethnic or other categories. These include the exclusion of: (i) cell phone-only households (which tend to be of lower socioeconomic status); (ii) respondents for whom Spanish and English is not their primary spoken language (and which could result in inflated or depressed estimates for some racial/ethnic groups); and (iii) military and institutionalized populations. Smaller sample sizes for some racial/ethnic groups, such as American Indian or Alaska Natives, also hinder the precision of their estimates of smoking prevalence. Moreover, ethnic categories used in national surveys (e.g., Asian) may not capture the heterogeneity present among various groups within that category. This is illustrated by data from California, which collects more detailed information on behavioral risk factors among its large Asian populations; for example, whereas the prevalence of smoking among Japanese American females (15.6%) is similar to that for non-Hispanic Whites (15.9%), the rate for Chinese American females is substantially lower (2.2%) ("Vital Signs," 2010). Underreporting of unhealthful behaviors such as smoking may also occur for some ethnic groups, depending on the sampling methodology used. Specifically, there is evidence that the prevalence of tobacco smoking is substantially higher among African Americans when random household sampling using same-race interviewers and neighborhood factors are considered; rates of smoking as high as 59% among Black men in high-poverty/high-segregated neighborhoods have been observed using these methods.


Tobacco Smoking and Cancer Risk

Tobacco use is a major cause of disease and premature death in the United States and many other countries (Centers for Disease Control and Prevention, 2008). Each year, more than 150,000 smokers in the United States die from smoking-attributable cancer, with about 80% of these deaths from lung cancer. Smoking also accounts for $193 billion in health care expenditures and productivity losses (Centers for Disease Control and Prevention, 2008).

Tobacco smoking is causally related to at least 16 types of cancers (International Agency for Research on Cancer, 2004), including cancers of the lung, colon and rectum, oral cavity, nasal cavities and nasal sinuses, pharynx, larynx, esophagus (squamous cell carcinoma and adenocarcinoma), stomach, pancreas, liver, urinary bladder, kidney (adenocarcinoma and transitional cell carcinoma), and the uterine cervix, and myeloid leukemia. Among these, the strongest association is with lung cancer (U.S. Department of Health and Human Services, 2004). Environmental tobacco smoke is responsible for an additional 3,000 lung cancer deaths among nonsmokers (Centers for Disease Control and Prevention, 2002). The health risks of tobacco use are not limited to cigarette smoking. Cigar, pipe, and smokeless tobacco use also increase the risk of cancer (U.S. Department of Health and Human Services, 2004). In addition to cancer, smoking is linked with the risk of other chronic conditions such as cardiovascular disease (Eyre et al., 2004).

Disparities in Prevalence of Tobacco Smoking

Over the past four decades, public health efforts have led to enormous reductions in tobacco use in the United States (Centers for Disease Control and Prevention, 2008). Still, 1 in 5 adults in the United States continues to smoke, with some subgroups smoking at even greater levels. An estimated 20.9% of U.S. adults currently smoke, but rates are higher among men (23.2%) than among women (18.1%). Although the prevalence of adult cigarette smoking varies by gender, as well as race or ethnic group, it is important to note age-group differences within these categories (see Table 3.1). For example, in considering race/ethnicity alone, smoking prevalence is similar for White and African American males (25% and 23.5%, respectively). Among White males, however, 31- to 44-year-olds smoke at higher rates (31.6%) compared to males 45 years or older (20.0%). In contrast, for African American males, the rates are more similar for the two age groups (21.7% and 25.5%) (see Table 3.1).

The prevalence of smoking also varies by level of educational attainment, with the highest prevalence of cigarette smoking among individuals who have only attended or completed high school. Also, men and women whose income is less than twice the poverty level are much more likely to be current smokers compared to those with higher incomes (see Table 3.1). Regional differences in tobacco use have also been noted, with smoking prevalence generally higher in the Southern states, including traditionally tobacco-growing states, and the Midwest (Jemal et al., 2008). These geographic differences have been correlated with less stringent tobacco control laws and policies (Jemal, Cokkinides, Shafey, & Thun, 2003).

TABLE 3.1 The Proportion of Persons Who Were Current Smokers by Race/Ethnicity, Age, Education Level, and Income Level, Adults 18 and Older, United States

% Current Smokersa




Race/ethnicity by age










45 years or older



African American









45 years or older





9. 5







45 years or older


9. 1

Asian American









45 years or older



American Indian and Alaska Nativeb









45 years or older




Below poverty level



At 100% to 200% above poverty level



>200% of poverty level



Education (years)d

8 or fewer






12 or GED



1 3-1 5






More than 16



Percentages are age adjusted to the 2000 U.S. standard population.

aSmoking is defined as persons who reported having smoked at least 100 cigarettes or more and who reported now smoking every day or some days.

bAmerican Indian/Alaska Native only, the estimated smoking prevalence should be interpreted with caution because of the small sample sizes.

cFamily income estimates are based on reported family income and poverty thresholds published by the U.S. Census Bureau.

dPrevalence by education level is for those >25 years.

Reducing Initiation and Prevalence of Tobacco Smoking

Tobacco control policies and laws, along with educational and clinical strategies—all with a solid evidence base—can prevent the initiation of smoking or help smokers quit. As is the case for cardiovascular disease, many of the risks associated with tobacco use decrease after cessation. Interventions that encourage quit attempts or assist smokers in quitting thus represent the surest short-term strategies to reduce lung cancer mortality at the population level (Curry et al., 2003; Doll, Peto, Boreham, & Sutherland, 2004; U.S. Department of Health and Human Services, 2000). To sustainably reduce the level of tobacco consumption and diseases in the long term, however, population-level strategies to prevent the initiation of smoking among youth are needed. This is most effectively achieved by strong advocacy that leads to legislative changes. Policies with demonstrated effects in reducing the initiation into smoking include limiting youth access and marketing of tobacco products, legislating smoke-free environments, increasing tobacco taxes, and conducting antismoking campaigns (Bala, Strzeszynski, & Cahill, 2008). In addition, because of the tobacco industry’s predatory practices in encouraging youth initiation, monitoring of their new products and promotions, and the heft of the law to curtail them, are needed. The new Family Smoking Prevention and Tobacco Control Act enacted in 2009 is a step in this direction.

Advocacy for and the implementation of legislation at the population level is also crucial to help reduce tobacco consumption by helping smokers quit. Health care reform that includes reimbursement for cessation treatments, or increasing taxes on cigarettes, strongly motivates many smokers to quit. For example, in anticipation of cigarette tax increases, thousands of smokers in the state of Michigan inundated the state-sponsored telephone counseling quit line, which led to its closure for several weeks in 2009 ("Quit-Smoking Hotline," March, 2009). This example also illustrates that once smokers are motivated to quit, individual-level interventions, which are characterized by higher levels of personal interaction between the targets of the interventions and their providers, are needed to provide smokers with the tools to help them do so successfully. Individual-level interventions are based on psychosocial or biomedical explanations for behavior, and include cognitive-behavioral strategies implemented in a variety of settings, as well as pharmacologic treatment. These methods can effectively double or triple smokers’ ability to quit, and thus play an important role in reducing the prevalence of smoking.

Reducing Disparities in Initiation and Prevalence of Tobacco Smoking

Although we know what interventions effectively prevent or reduce smoking prevalence, more research is needed to investigate the extent to which they reduce or eliminate disparities in tobacco consumption and cessation.

Reducing the burden of tobacco for all individuals, regardless of minority status, is an overarching goal (Zhu, Hebert, Wong, Cummins, & Gamst, 2010) and so investigators have typically not geared population-level interventions to reduce disparities in tobacco use or cessation. Nevertheless, researchers have begun reporting the results of more in-depth analyses that determine whether the effects of broad-based, population-level interventions have similar or pronounced effects among certain sociodemographic or economically vulnerable groups. Evidence that population-level strategies not targeted to specific populations can nevertheless disproportionately impact them is illustrated by the finding that the substantial gap in lung cancer death rates between Black and White young adults (20-39 years) in the early 1990s was virtually eliminated by the mid-2000s, a result attributed to cigarette price increases and other intervention strategies that strongly impacted Black young adults (Jemal, Center, & Ward, 2009).

The lethal effect of disparities in tobacco use (e.g., higher lung cancer rates in older African American smokers) has also encouraged investigators to examine whether individual-level interventions that acknowledge or incorporate the differing experiences, attitudes, and knowledge of a particular group will be more powerful—and thus more successful—in achieving behavior change compared to interventions that do not. An example of such cultural adaptation would be a counseling program to help African American smokers quit that includes strategies to help them better cope with stressors resulting from racial discrimination, a possible trigger of urges to smoke and relapse.

In the sections that follow, we first review population-level interventions that, while not necessarily targeted toward vulnerable populations, nonetheless have been evaluated for their ability to reduce disparities in tobacco use. We then review the evidence for the effectiveness of individual-level smoking cessation interventions among populations experiencing disparities in tobacco use and cessation. We briefly describe how these treatments have been modified or expanded in attempts to increase the likelihood of cessation among these populations, and we discuss their impact, if any, in changing disparities in tobacco use or cessation.

Effects of Population-Level Interventions in Reducing Disparities

Some interventions with the largest impact have been at the population level, which recognizes that behavior change is a result of social and environmental influences. As noted earlier, the reduction in tobacco prevalence from its high of 64% to its current 20% of the population can be attributed to a number of population-level strategies that effectively reduce the likelihood of smoking initiation. For example, higher tobacco taxes make the cost of cigarettes prohibitive for youth and have been shown to be associated with lower uptake of smoking in this population (Levy, Cummings, & Hyland, 2000a, 2000b). Tobacco taxes thus represent an effective primary prevention tool to reduce smoking consumption in the long term. Young Black adults may be particularly sensitive to these increases, which could explain the recent convergence in lung cancer rates between Black and White young adults (Jemal et al., 2009). In general, however, data is lacking on whether higher tobacco taxes prevent initiation to a greater or lesser degree among youth belonging to other groups that have traditionally experienced disparities.

Higher cigarette prices also lead to reductions in prevalence by encouraging existing smokers to quit (Farrelly, 2001; Farrelly, Pechacek, Thomas, & Nelson, 2008). Studies have now also documented that cigarette price increases have larger impacts on socioeconomically vulnerable populations, such as those from lower-income brackets and minority populations (Farrelly, 2001; Siahpush, Wakefield, Spittal, Durkin, & Scollo, 2009). These effects are reliable and have been reported in samples from Canada, the United Kingdom (U.K.), the United States, and other countries. A particularly well-designed study in Australia using monthly (rather than yearly) data on smoking prevalence and cigarette prices (because such data allow for greater sensitivity in detecting causal associations) found that lower-income groups were most responsive to increases in price, as evidenced by their greatest reduction in prevalence. The effect of tobacco taxes in reducing disparities can be striking. For example, Farrelly et al. found that compared to Whites, responsivity to price increases was two times greater for African Americans, and six times greater for Hispanics (Farrelly, 2001).

Antitobacco-advertising campaigns also play a role in preventing initiation into smoking (Emery et al., 2005; Farrelly, Davis, Haviland, Healton, & Messeri, 2005; Siegel & Biener, 2000), although the extent to which they limit initiation among members of special populations has not been investigated. There is evidence, however, that antitobacco campaigns have the potential to reduce disparities in the likelihood of cessation, but that the content of antitobacco-advertising messages appears to be a crucial factor. In a longitudinal study that examined the effects of 134 different antitobacco television advertisements broadcast in Massachusetts over a 2-year period, advertisements that were "highly emotional," where the narrator described the negative health effects of smoking on oneself or family members, or that included personal testimonials, had the strongest effects on quitting among low- and middle-SES (socioeconomic status) groups. In contrast, for high-income individuals (at least some college and >$50,000 per year), there were no differences between the type of advertisements and the likelihood of quitting (Durkin, Biener, & Wakefield, 2009). Differential effects of the advertisements as a function of ethnicity or race were not reported, however.

Considering that African American communities are disproportionately targeted by tobacco industry advertising (Primack, Bost, Land, & Fine, 2007), it will be imperative that counter-advertising has similar or stronger effects on African American smokers if disparities in tobacco use and its effects are to be eliminated for this group.

Effects of Community-Level Interventions in Reducing Disparities

Community-level interventions aimed at preventing smoking initiation among children and adolescents are typically school based. Multicomponent programs that include education combined with learning to resist social influences to smoke, deconstructing media influences promoting tobacco use, skills training, and helping to construct and deliver the intervention have shown positive short-term effects (i.e., between 1 and 3 years) (Dobbins, DeCorby, Manske, & Goldblatt, 2008). Effects are even greater when 15 or more sessions up to the ninth grade are delivered (Flay, 2009), and when combined with media and policy interventions such as smoking bans and increased taxation (Dobbins et al., 2008). Data is lacking, however, on whether the efficacy of school-based prevention programs differ as a function of gender, racial/ethnic group, or other sociodemographic variables (Dobbins et al., 2008).

For individuals who already smoke, one of the most popular community-level interventions is the Quit and Win contests held in many communities in North America, and now internationally in about 80 countries. The contests originated from the Minnesota Heart Health Program, are publicized through mass media outlets, and attempt to engage smokers from entire communities by offering incentives and prizes to those who successfully quit (Cahill & Perera, 2008). In one randomized controlled trial of a Quit and Win contest, women and lower-income smokers, who typically have lower quit rates compared with men and higher-income smokers, were equally likely to successfully stop smoking (Hahn et al., 2005). The homogeneity of participants in many Quit and Win contests, however (i.e., predominantly White, younger, greater educational attainment) (Cahill & Perera, 2008), makes it difficult to determine whether special populations are, in general, equally likely to benefit from them. Other methodological concerns have plagued evaluations of the effects of community interventions other than Quit and Win in reducing smoking prevalence. For example, in a review of community interventions in general, Secker-Walker and colleagues pointed out that many studies were underpowered, did not use randomization to interventions, or used the individual as the unit of analysis when the community was the unit of assignment (Secker-Walker, Gnich, Platt, & Lancaster, 2002). In the most methodologically rigorous study of the set of studies that they examined (the COMMIT study), however, men appeared to benefit somewhat more than women, and lighter smokers more than heavier smokers (Fisher, 1995).

Next post:

Previous post: