Abstract
The incidence of colorectal cancer (CRC) among Hispanics in the state of New Mexico has increased in the past decade while that among whites has declined significantly. Using the 2006 New Mexico Behavioral Risk Factor Sur- veillance System (BRFSS) survey, we compared CRC screening among Hispanics and whites by gender to examine the influence of demographic, socioeconomic, preventive health, and clinical measures on the utilization of CRC screening. Although we found no ethnic differences in the prevalence of current breast, cervical and cancer screening, Hispanics were less likely to be current with CRC screening than whites. These differences wereobserved acrossa range of socioeconomic and other explanatory measures and in both genders. Hispanics also had a higher prevalence of CRC- related risk factors than whites, including inactivity, obesity, and diabetes, and ranked lower for most socioeconomic measures. Adjusting for healthcare coverage, education, and income in logistic regression models eliminated the Hispanic- white differences in CRC screening among men, and sub- stantially reduced but did not eliminate screening differences among women. Innovative methods are needed to reach Hispanics to raise awareness of and participation in CRC screening. Because many CRC risk factors are potentially modifiable, appropriate cultural and linguistic interventions tailored to specific Hispanic subgroups and aimed at pro- moting CRC screening and reducing CRC risk factors may decrease ethnic disparities in CRC incidence.
Keywords: Behavioral risk factor surveillance system, Hispanics, Healthcare disparities, Colorectal cancer, Screening
Introduction
Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the third leading cause of cancer death in the United States in both men and women. Nationally, colorectal cancer incidence and mortality rates have declined for men and women and for all racial/ethnic groups over the past decade [1]. However, in New Mexico, the incidence of CRC among Hispanics has continued to increase in the last decade while incidence among whites has declined significantly [2].
Reasons for the ethnic differences in CRC incidence are unclear. A number of explanations have been proposed based on national data related to screening uptake and risk factors. Screening is recommended for average-risk adults beginning at age 50. Fecal occult blood testing [FOBT] and lower endoscopy are proven to reduce CRC incidence and are highly recommended screening tests [3]. Though CRC screening rates have increased in the US, Hispanic men and women are less likely to comply with screening guidelines than whites at both the state and national levels [4–7]. Combined data from the 2004 and 2006 New Mexico Behavioral Risk Factor Surveillance System (BRFSS) surveys showed that 45.7 % of Hispanic respondents, and 57.6% of white respondents were currently screened for CRC based on having had either a FOBT in the previous year and/or a lower endoscopy within the previous ten years [3]. English-language proficiency and country of origin have also been shown to influence CRC screening uptake among Hispanics [8]. Hispanics also tend to have a higher prevalence of CRC risk factors than whites, including physical inactivity, overweight or obesity, and diabetes, though the prevalence of these factors vary among Hispanic subgroups [9]. Explaining the disparities in CRC incidence in New Mexico is difficult because to date little is known about differences in CRC screening patterns and CRC risk factors between Hispanics and whites.
We used data from the 2006 New Mexico Behavioral Risk Factor Surveillance System (BRFSS) survey to (a) compare the prevalence of CRC screening, and demographic, socio- economic, preventive health, and clinical measures in His- panics and whites by gender, and (b) to examine the influence of these measures on the utilization of CRC screening.
Methods
BRFSS Data
Data for the current analysis were obtained from the 2006 BRFSS survey of 3303 New Mexico respondents 50 years of age and older [10]. The BRFSS is a nationwide, random- sample, telephone survey that has collected data on health- related behaviors, preventive health practices, and access to health care among the non-institutionalized, civilian US population age 18 years and older since 1994. The study participants were a population-based sample of respondents to the 2006 New Mexico BRFSS, a federally funded annual survey implemented in collaboration with state health departments. Random-digit dialing methods were used to derive a probability sample of households with telephones to collect data on health-related behaviors and risk factors for respondents aged 18 years or older. Survey respondents who identified themselves as Hispanic or Latino were categorized as such regardless of race. In New Mexico, the predominant Hispanic subgroups are Mexican, Mexican- American and Spanish American, whose ancestors became US citizens following annexation in 1848.
Measures
Every two years, respondents aged 50 years and older are asked questions about CRC screening by fecal occult blood test (FOBT), or sigmoidoscopy/colonoscopy (lower endoscopy), and when these tests were last performed. To be consistent with previous BRFSS reports and screening guidelines, we define current CRC screening as having had an FOBT within the past year and/or lower endoscopy within the past 10 years [7].
The demographic measures collected in the BRFSS included age, marital and veteran status and the language in which the survey was administered, either Spanish or English. Socioeconomic measures include health care coverage, cur- rent employment status, annual household income, and high school-level education. Veteran status was only included as an indicator for health care coverage among men due to the low numbers of female veterans in the survey population. The clinical measures included perceived general health, self- reported diabetes, and self-reported height and weight (used to calculate body mass index (BMI). Lifestyle measures included smoking, risk for heavy alcohol consumption, and physical activity. Preventive health measures include screening for breast, cervical, prostate, and colorectal cancer and annual influenza vaccination. Female respondents who reported having a Pap smear or mammogram in the past 12 months would be considered currently adhering to cervical and breast cancer screening guidelines. Male respondents who had a prostate-specific antigen (PSA) test in the past 24 months were considered current for prostate cancer screening.
Statistical Analysis
We report weighted population-based proportions with confidence intervals (CI) to describe the distribution of CRC screening and other measures (Table 1). P values for Chi square analyses are reported for prevalence estimates stratified by gender and comparing Hispanics to whites within the gender groups. Differences in CRC screening prevalence between Hispanics and whites, stratified by sex, and adjusted by individual survey measures are shown in Table 2. Differences within gender and ethnicity groups were evaluated by comparing confidence intervals. Table 3 shows the ‘‘risk’’ of Hispanics being current with CRC screening as compared to whites, stratified by sex, and adjusted for demographic, socioeconomic, clinical, life- style and preventive health factors. Odds ratios (OR) were estimated by multivariate logistic regression sequentially adjusted for groups of demographic, socioeconomic, clinical, lifestyle, and preventive measures, and all measures combined. Interactions between the individual study measures and Hispanic versus white ethnicity, and survey language were tested. The data were analyzed using SUDAAN 10.1 (RTI, Research Triangle Park, NC).
Table 1.
Distribution of colorectal cancer risk factors (demographic, socioeconomic, clinical and health maintenance) reported by Hispanic and white New Mexicans age 50 years and over, stratified by sex and Hispanic/white ethnicity: data from the New Mexico 2006 BRFSS survey
| Males
|
Females
|
|||||
|---|---|---|---|---|---|---|
| Hispanic Percent (95% confidence interval) | White Percent (95% confidence interval) | P valuea | Hispanic Percent (95% confidence interval) | White Percent (95% confidence interval) | P valuea | |
| Demographic factors | ||||||
| Age | ||||||
| 50–60 years | 53.7 (47.6–59.7) | 47.2 (43.4–51.0) | .03 | 49.2 (44.4–45.1) | 40.8 (37.7–44.0) | <001 |
| 61–70 years | 26.8 (21.8–32.5) | 25.7 (22.5–29.2) | 29.5 (25.2–34.2) | 27.5 (24.9–30.2) | ||
| >70 years | 19.4 (15.2–24.5) | 27.1 (23.8–30.6) | 21.2 (17.7–25.2) | 31.7 (28.9–34.6) | ||
| Survey language | ||||||
| English | 80.1 (74.3–84.8) | 99.9 (99.6–99.9) | <.001 | 85.3 (81.6–88.2) | 100.0 | <.001 |
| Spanish | 19.9 (15.2–25.6) | 0.1 (0.1–0.4) | 14.8 (11.7–18.4) | 0.0 | ||
| Marital status | ||||||
| Married, unmarried couple | 77.4 (72.3–81.1) | 76.9 (73.6–79.4) | .86 | 56.7 (51.9–61.3) | 58.6 (55.5–61.4) | .49 |
| Divorced, widowed, separated never married | 22.6 (18.4–27.1) | 23.1 (20.3–26.0) | 43.3 (38.7–48.0) | 41.4 (38.3–44.3) | ||
| Live in or near a metropolitan area | ||||||
| Yes | 60.0 (54.3–65.4) | 64.1 (60.9–67.1) | .23 | 62.1 (57.9–66.1) | 63.7 (61.2–66.0) | .56 |
| No | 40.0 (34.5–45.7) | 35.9 (32.8–39.1) | 37.9 (33.8–42.1) | 36.4 (34.0–38.7) | ||
| Armed services veteran | 44.0 (37.9–50.2) | 52.7 (48.9–56.5) | .02 | d | ||
| Socioeconomic factors | ||||||
| Education | ||||||
| <High school | 27.1 (21.9–32.8) | 5.8 (4.3–7.7) | <.001 | 32.4 (28.1–37.0) | 5.8 (4.5–7.2) | <.001 |
| ≥High school | 72.9 (67.0–77.9) | 94.2 (92.2–95.6) | 67.5 62.8–71.8) | 94.3 (92.4–95.2) | ||
| Annual household income | ||||||
| <$25,000 | 46.7 (40.4–53.2) | 17.4 (14.8–20.4) | <.001 | 55.3 (50.0–60.5) | 28.1 (25.4–31.0) | <.001 |
| $25,000–$50,000 | 27.7 (22.3–33.8) | 31.2 (27.6–34.9) | 27.5 (22.9–32.5) | 30.0 (27.1–33.2) | ||
| >$50,000 | 25.5 (20.1–31.8) | 51.4 (47.4–55.3) | 17.2 (13.6–21.5) | 41.8 (38.5–45.2) | ||
| Employment | ||||||
| Employed | 47.4 (41.2–53.7) | 51.1 (47.2–54.9) | .006 | 38.0 (33.4–42.8) | 37.4 (34.4–40.5) | .001 |
| Retired | 36.1 (30.5–42.1) | 40.4 (36.7–44.1) | 30.9 (26.8–35.5) | 39.7 (36.8–42.8) | ||
| Not employed | 16.5 (12.6–21.3) | 8.6 (6.6–11.1) | 31.1 (26.7–35.8) | 22.8 (20.3–25.5) | ||
| Health insurance | 81.0 (74.8–84.6) | 90.9 (88.5–92.8) | <.001 | 80.9 (76.4–84.1) | 93.3 (91.5–94.6) | <.001 |
| Clinical factors | ||||||
| BMI | ||||||
| ≤25 | 24.5 (18.7–28.8) | 34.4 (30.6–37.9) | 0.006 | 32.9 (26.7–35.8) | 47.4 (42.7–49.0) | <.001 |
| 26–30 | 49.5 (41.1–53.6) | 46.3 (42.2–49.8) | 39.8 (33.0–42.3) | 32.9 (28.9–34.8) | ||
| >30 | 26.0 (19.8–30.6) | 19.3 (16.3–22.4) | 27.2 (21.7–29.9) | 19.7 (16.7–21.7) | ||
| Diabetes | 21.1 (16.4–26.6) | 19.3 (7.1–11.7) | <.001 | 19.1 (15.6–23.1) | 9.8 (8.0–11.8) | <.001 |
| Perceived general health | ||||||
| Excellent | 16.3 (12.0–21.7) | 21.6 (18.5–24.9) | <.001 | 13.3 (10.2–17.1) | 20.7 (18.2–23.3) | <.001 |
| Very good/good | 46.3 (40.1–52.4) | 59.8 (55.9–63.4) | 49.4 (44.4–54.2) | 60.9 (57.7–63.7) | ||
| Fair/poor | 37.4 (31.6–43.4) | 18.6 (15.7–21.7) | 37.3 (32.5–42.0) | 18.4 (16.2–20.7) | ||
| Lifestyle factors | ||||||
| Current smoker | 20.2 (15.8–25.3) | 15.5 (12.8–18.4) | .09 | 18.4 (14.6–22.7) | 13.7 (11.8–15.7) | .04 |
| At risk for heavy alcohol consumptionb | 4.9 (2.9–7.7) | 5.2 (3.7–6.9) | .85 | 1.6 (0.6–3.9) | 4.2 (3.1–5.6) | .008 |
| Physical activity or exercise in past 30 daysc | 69.5 (63.6–74.9) | 78.7 (75.3–81.7) | .006 | 64.6 (59.8–69.2) | 75.0 (71.9–77.3) | <.001 |
| Health behaviors | ||||||
| Influenza vaccine ≤12 months ago | 44.8 (38.6–51.1) | 50.2 (46.3–54.0) | .16 | 45.2 (40.4–50.1) | 51.7 (48.5–4.8) | .03 |
| CRC screeningd | ||||||
| Current | 45.3 (39.2–51.6) | 56.1 (52.3–60.0) | <.001 | 42.5 (37.7–45.4) | 57.9 (54.8–60.9) | <.001 |
| Not current | 6.0 (3.8–9.3) | 10.7 (8.5–13.3) | 7.6 (5.5–10.5) | 12.9 (10.9–15.2) | ||
| Never | 48.7 (42.5–54.9) | 33.2 (29.7–36.9) | 49.9 (45.0–54.9) | 29.2 (26.4–32.1) | ||
| FOBT | ||||||
| ≤12 months ago | 13.0 (9.29–18.0) | 12.6 (10.3–15.3) | <.001 | 11.0 (8.42–14.2) | 13.0 (11.0–15.3) | <.001 |
| >12 months ago | 14.4 (10.7–19.7) | 26.5 (23.1–30.1) | 17.9 (14.4–22.0) | 32.4 (29.4–35.4 | ||
| Never | ||||||
| Lower endoscopy | ||||||
| ≤10 years ago | 39.4 (33.4–45.6) | 51.3 (47.4–55.1) | .003 | 38.5 (33.9–43.3) | 53.0 (49.9–56.2) | <.001 |
| >10 years ago | 2.3 (1.3–5.9) | 3.8 (2.7–5.3) | 2.0 (1.1–3.6) | 5.0 (3.8–6.6) | ||
| Never | 57.8 (51.6–36.9) | 44.5 (40.8–48.4) | 59.3 (54.5–64.0) | 41.8 (38.8–44.9) | ||
| PSA test | ||||||
| ≤24 months ago | 76.4 (41.7–54.3) | 77.2 (57.9–65.4) | .13 | e | ||
| >24 months ago | 13.2 (5.3–12.7) | 16.6 (10.7–16.4) | ||||
| Never | 10.4 (4.4–9.6) | 6.2 (3.6–6.8) | ||||
| PAP test | f | |||||
| ≤12 months ago | 39.4 (34.6–44.3) | 39.0 (36.0–42.2) | .18 | |||
| >12 months ago | 15.1 (11.8–19.1) | 16.6 (14.3–19.1) | ||||
| Never | 4.1 (2.6–6.2) | 1.9 (1.2–3.0) | ||||
| Had hysterectomy | 41.5 (36.6–46.5) | 42.4 (39.3–45.6) | ||||
| Mammogram | f | |||||
| ≤12 months ago | 53.9 (48.8–58.8) | 57.3 (54.1–60.5) | .10 | |||
| >12 months ago | 37.8 (33.0–42.7) | 37.5 (34.4–40.7) | ||||
| Never | 8.4 (6.0–11.0) | 5.15 (4.0–6.5) | ||||
BMI body mass index, CRC colorectal cancer, FOBT fecal occult blood test, PSA prostate-specific antigen, PAP Papanicolaou
p-value for Chi-square test for difference between racial/ethnic groups
Having more than two drinks per day for adult men; more than one drink per day for adult women
Physical activity and/or exercise outside of a regular job
Not reported in women due to low frequency
Not applicable to women
Not applicable to men
Table 2.
Current colorectal cancer screening by selected demographic, socioeconomic, clinical and health maintenance factors, stratified by sex, as reported by Hispanic and white New Mexicans age 50 years and over in the 2006 New Mexico Behavior Risk Factor Surveillance Survey
| Males
|
Females
|
|||||
|---|---|---|---|---|---|---|
| Hispanic Percent (95% confidence interval) | White Percent (95% confidence interval) | P valuea | Hispanic Percent (95% confidence interval) | White Percent (95% confidence interval) | P valuea | |
| Demographic factors | ||||||
| Age | ||||||
| 50–60 years | 40.3 (31.9–49.3) | 46.7 (41.0–52.5) | <.001 | 37.2 (30.5–44.4) | 49.0 (43.8–54.2) | .11 |
| 61–70 years | 47.8 (36.6–59.2) | 64.1 (56.8–70.9) | 48.7 (39.7–57.7) | 66.0 (60.7–71.0) | ||
| >70 years | 55.8 (43.0–67.9) | 64.9 (57.5–71.7) | 46.2 (36.8–55.8) | 62.3 (57.0–67.4) | ||
| Survey language | ||||||
| English | 47.9 (41.2–54.7) | 56.1 (52.2–59.9) | .02 | 44.8 (39.5–50.1) | 57.9 (54.8–61.0) | <.001 |
| Spanish | 35.0 (21.6–51.3) | <.10 | 29.1 (18.9–42.0) | 0.0 | ||
| Marital status | ||||||
| Married, unmarried couple | 47.7 (40.4–55.1) | 59.7 (55.1–64.1) | <.001 | 45.0 (38.4–51.8) | 61.3 (56.8–65.5) | .005 |
| Divorced, widowed, separated never married | 38.2 (29.0–48.4) | 43.9 (37.4–50.6) | 39.1 (32.6–46.1) | 53.4 (49.1–57.5) | ||
| Live in or near a metropolitan area | ||||||
| Yes | 35.6 (27.5–44.5) | 56.0 (50.8–61.2) | .10 | 42.8 (36.1–49.7) | 59.4 (55.1–63.6) | .21 |
| No | 51.9 (43.4–60.3) | 56.4 (51.1–61.5) | 42.0 (35.9–48.3) | 55.2 (51.0–59.3) | ||
| Armed services veteran | d | |||||
| Yes | 53.9 (42.1–61.0) | 62.3 (53.0–63.4) | <.001 | |||
| No | 39.7 (31.7–48.2) | 49.9 (44.3–55.5) | ||||
| Socioeconomic factors | ||||||
| Education | ||||||
| <High school | 31.9 (21.8–44.1) | 35.1 (22.4–50.2) | <.001 | 33.9 (26.8–41.9) | 46.0 (34.6–57.9) | <.001 |
| ≥High school | 50.4 (43.2–57.7) | 57.5 (53.5–61.4) | 46.7 (40.7–52.8) | 58.7 (55.5–61.9) | ||
| Annual household income | ||||||
| <$25,000 | 36.5 (28.0–45.9) | 45.3 (36.9–53.9) | .001 | 36.7 (30.4–43.4) | 49.3 (43.6–55.0) | <.001 |
| $25,000–$50,000 | 53.4 (41.4–65.0) | 50.3 (43.3–57.2) | 41.6 (32.0–51.9) | 61.2 (55.1–67.0) | ||
| >$50,000 | 59.8 (46.1–72.1) | 63.1 (57.2–68.6) | 54.1 (41.5–66.1) | 61.1 (55.4–66.6) | ||
| Employment status | ||||||
| Employed | 40.2 (31.1–50.1) | 51.6 (46.0–57.1) | <.001 | 35.2 (28.1–43.0) | 53.0 (47.6–58.5) | <.001 |
| Retired | 57.7 (48.1–66.8) | 64.5 (58.7–70.0) | 51.9 (43.6–60.2) | 64.3 (59.7–68.6) | ||
| Not employed | 33.4 (21.8–47.5) | 45.2 (32.2–58.9) | 33.4 (21.8–47.5) | 45.2 (32.2–58.9) | ||
| Health insurance | ||||||
| Yes | 47.1 (40.3–54.0) | 58.9 (54.8–62.9) | <.001 | 46.1 (40.9–51.5) | 60.7 (57.5–63.9) | <.001 |
| No | 36.6 (23.4–52.1) | 28.6 (18.3–41.6) | 27.2 (17.2–40.2) | 18.9 (11.0–30.5) | ||
| Clinical factors | ||||||
| BMI | ||||||
| ≤25 | 44.7 (32.9–57.0) | 55.0 (48.6–61.3) | .91 | 47.3 (38.4–56.4) | 57.3 (52.7–61.8) | .64 |
| 26–30 | 44.1 (35.2–53.4) | 56.8 (51.0–62.4) | 41.6 (33.8–49.8) | 59.4 (53.9–64.7) | ||
| >30 | 49.4 (37.4–61.6) | 57.5 (48.2–66.2) | 38.4 (30.1–47.4) | 60.8 (53.7–67.4) | ||
| Diabetes | ||||||
| Yes | 56.4 (43.0–69.0) | 60.7 (47.2–72.6) | .20 | 56.0 (45.2–66.2) | 57.8 (47.5–67.5) | .23 |
| No | 42.4 (35.6–49.4) | 55.7 (51.6–59.7) | 39.3 (34.1–44.8) | 57.9 (54.6–61.1) | ||
| Perceived general health | ||||||
| Excellent | 45.8 (30.1–62.4) | 55.0 (46.6–63.1) | .06 | 39.5 (27.0–53.4) | 57.7 (50.8–64.4) | .98 |
| Very good/good | 49.4 (40.6–58.3) | 58.2 (53.2–62.9) | 43.7 (37.1–50.6) | 56.2 (52.1–60.2) | ||
| Fair/poor | 40.4 (31.0–50.4) | 51.6 (42.6–60.6) | 42.1 (34.6–50.1) | 63.7 (56.9–69.9) | ||
| Lifestyle factors | ||||||
| Current smoker | ||||||
| Yes | 40.0 (27.8–53.7) | 31.8 (23.4–41.7) | <.01 | 22.8 (14.7–33.8) | 43.6 (36.2–51.3) | <.001 |
| No | 46.4 (39.4–53.4) | 60.6 (56.4–64.6) | 46.9 (41.6–52.2) | 60.3 (56.9–63.6) | ||
| At risk for heavy alcohol consumptionb | ||||||
| Yes | 31.4 (14.7–54.9) | 55.8 (39.7–7 0.7) | .47 | 35.9 (9.5–74.8) | 68.9 (53.4–81.1) | .17 |
| No | 47.6 (41.2–54.2) | 56.6 (52.6–60.5) | 43.1 (38.2–48.1) | 58.2 (54.9–61.3) | ||
| Physical activity or exercise in past 30 daysc | ||||||
| Yes | 46.4 (39.0–54.0) | 57.4 (53.1–61.6) | .13 | 44.9 (38.9–51.0) | 60.0 (56.4–63.5) | .002 |
| No | 43.0 (32.4–54.3) | 51.6 (43.2–60.0) | 38.1 (30.3,46.5) | 51.6 (45.4–57.7) | ||
| Health behaviors | ||||||
| Influenza vaccine | ||||||
| ≤12 months ago | 57.5 (48.5–66.1) | 69.5 (64.1–74.4) | <.001 | 50.4 (43.4–57.4) | 69.6 (65.6–73.4) | .001 |
| >12 months ago | 36.7 (28.7–45.4) | 43.3 (38.0–48.9) | 37.8 (31.3–44.7) | 46.9 (42.3–51.5) | ||
| PSA test | e | |||||
| ≤24 months ago | 69.3 (60.8–76.7) | 69.3 (64.5–73.7) | <.001 | |||
| >24 months ago | 55.0 (32.5–75.5) | 47.1 (35.0–58.5) | ||||
| Never | 18.9 (11.8–29.1) | 35.3 (20.5–35.2) | ||||
| PAP test | f | |||||
| ≤12 months ago | 39.5 (32.0–47.3) | 64.7 (59.8–69.3) | <.001 | |||
| >12 months ago | 31.0 (27.6–41.2) | 33.5 (41.1–62.7) | ||||
| Never | 11.2 (3.9– 28.4) | 11.1 (3.6– 29.3) | ||||
| Had hysterectomy | 57.5 (49.6–65.0) | 67.8 (62.0–72.3) | ||||
| Mammogram | f | |||||
| ≤12 months ago | 54.2 (47.3–61.0) | 73.9 (70.2–77.3) | <.001 | |||
| >12 months ago | 37.2 (29.5–45.6) | 44.7 (39.4–50.0) | ||||
| Never | 16.9 (8.2–31.7) | 12.5 (6.47–22.2) | ||||
BMI body mass index, CRC colorectal cancer, FOBT fecal occult blood test, PSA prostate-specific antigen, PAP Papanicolaou
p-value for Chi-square test for difference between racial/ethnic groups
Having more than two drinks per day for adult men; more than one drink per day for adult women
Physical activity and/or exercise outside of a regular job
Not reported in women due to low frequency
Not applicable to women
Not applicable to men
Table 3.
Odds ratios for current colorectal cancer screening in New Mexico Hispanics versus whites adjusted for demographic, socioeconomic, clinical, lifestyle and preventive health factors
| Modelsa | Men Odds ratio (95% confidence interval) |
Women Odds ratio (95% confidence interval) |
|---|---|---|
| Race/ethnicity | 0.69 (0.51–0.94) | 0.55 (0.44–0.70) |
| Race/ethnicity + survey languageb | 0.76 (0.55–1.05) | 0.61 (0.47–0.78) |
| Race/ethnicity + demographicc | 0.67 (0.50–0.91) | 0.56 (0.44–0.71) |
| Race/ethnicity + socioeconomicd | 1.09 (0.77–1.17) | 0.73 (0.56–0.97) |
| Race/ethnicity + clinicale | 0.70 (0.51–0.97) | 0.54 (0.42–0.69) |
| Race/ethnicity + lifestylef | 0.73 (0.53–0.99) | 0.58 (0.46–0.74) |
| Race/ethnicity + preventive healthg | 0.87 (0.63–1.21) | 0.56 (0.43–0.74) |
| Race/ethnicity + survey language + demographics + economic + clinical + lifestyle + preventive health | 1.26 (0.85–1.86) | 0.65 (0.48–0.90) |
Data from the 2006 New Mexico Behavior Risk Factor Surveillance Survey (BRFSS)
All models are adjusted for age, white is reference group
Survey Language includes Spanish and English
Demographic model includes age, marital status, urban versus rural residence, and veteran status for males
Socioeconomic model includes health care coverage, current employment status, annual household income and high school-level education e Clinical model includes perceived general health, obesity and diabetes
Lifestyle model includes smoking, heavy alcohol consumption, and physical activity
Preventive health model includes screening for breast, cervical, prostate and colorectal cancer and influenza vaccination
Results
Sample Characteristics
The 2006 New Mexico BRFSS data included 355 Hispanic men (mean age 61.5 years), 640 Hispanic women (mean age 62.4 years), 908 white men (mean age 63.7 years), and 1438 white women (mean age 65.2 years). Twenty percent Hispanic men and 15 % of Hispanic women were interviewed in Spanish. Compared to whites, Hispanics in this data set were younger, had less education and lower household income, were more likely to be smokers, less physically active, and perceive their general health to be fair or poor (Table 1). Regardless of gender, a larger proportion of Hispanic than white respondents reported being uninsured, having diabetes and a calculated BMI in the overweight to obese range, but a larger proportion of whites reported current for CRC screening. We noted no differences in CRC screening prevalence between men and women within the white and Hispanic groups based on overlapping confidence intervals. Hispanic men were more likely to have never been screened for prostate cancer or CRC than white men. Among those men who were current for CRC screening at the time of interview, screening by lower endoscopy was less common among Hispanics, but there were no ethnic difference in screening by FOBT. Hispanic women reported similar prevalence of current cervical and breast cancer screening as white women (P > 0.05), but lower prevalence of current CRC screening by either lower endoscopy or FOBT.
Current CRC Screening
When we examined current CRC screening adjusted for individual measures in men and women stratified by ethnicity (Table 2). The measures that were associated with being current for CRC across genders and ethnicities include older age, being married, retired, a non-smoker, an armed services veteran (males), and English speaker (Hispanics), being up-to-date with influenza vaccination and other cancer screening tests, and having a high school level education or greater. Several within ethnicity differences in CRC screening by gender are observed. White respondents, regardless of diabetic status, and Hispanic diabetics had similar prevalence of CRC screening, but the prevalence of CRC screening among Hispanic females who were non- diabetic was significantly lower. The proportion of CRC screening in men and white women increased with increasing BMI, but decreased at overweight and obese BMI among Hispanic women. Also, the difference in CRC screening between smokers and non smokers was larger among Hispanic women that white women.
Overall, the prevalence of CRC screening reported by Hispanic respondents was lower than the prevalence of CRC screening reported by white respondents who had the same characteristics. For example, approximately 47 % of Hispanics with health care coverage were up to date with CRC screening, compared to 60 % of whites with cover- age. Male smokers were an exception. In this group, white men were less likely to be screened than Hispanic men.
Adjusted CRC Screening Rates
In logistic regression models adjusted only for age, His- panic males were 31 % less likely to be currently screened for CRC (95% CI 0.51–0.94) than white males, and His- panic females were 45 % less likely (95% CI 0.44, 0.70) to have current screening than white females (Table 3). In subsequent models, adjusting for socioeconomic status, language preference, and preventive health factors reduced ethnic differences in CRC screening among males (95% CI included 1.00). Adjusting for socioeconomic status, His- panic women were 27 % less likely (95% CI 0.56, 0.97) to be up-to-date for CRC screening as compared to white women. Hispanic women were less likely to be current in their CRC screening when compared to white women according to all the logistic regression models. Lower odds for current CRC screening persisted between Hispanic and white women even in the model that adjusted for all risk factor measures simultaneously.
Discussion
The 2006 New Mexico BRFSS data show that Hispanic men and women were less likely to be current with recommended colorectal cancer screening than white men and women. Further, Hispanics also had a higher prevalence of CRC-related risk factors than whites, including inactivity, obesity, and diabetes. Ethnic differences in CRC screening were consistently observed across a range of socioeconomic, demographic, clinical, and health behavior factors. However, adjusting for socioeconomic factors essentially eliminated screening differences between Hispanic and white men, and substantially reduced screening differences between Hispanic and white women.
Our findings are consistent with previous reports on screening uptake. Even though CRC screening rates have increased since the launch of intervention programs targeting Hispanics, the increase has been less in Hispanics than in whites [5–7, 11]. CRC screening rates among Hispanics living in the western US have been significantly lower than rates in whites[12]. Similar to national studies, we noted no within-ethnicity gender differences in CRC screening, and Hispanic women are less likely to be screened for CRC than for breast or cervical cancer [11, 13].
Numerous socioeconomic-related barriers to CRC screening have been reported by Hispanics. These barriers range from cost of screening, lack of health insurance coverage or a usual source of health care to difficulty obtaining transportation to screening facilities [14]. Two factors, lack of insurance and not having a usual source of care, consistently explain Hispanic-white disparities in guideline-recommended preventive care services and colorectal cancer screening [14, 15]. In our logistic regression models, adjusting for health insurance coverage, income, education, and employment status eliminated screening differences between men, but did not completely account for the lower prevalence of CRC screening in Hispanic women. Hispanic women were less likely than white women to have health care coverage, and signifi- cantly more likely to report an annual household incomes of less than $25,000, having less than a high school level education, and being unemployed.
Given that socioeconomic status is associated with access to health care and with higher cancer screening rates, the overall prevalence of being up-to-date with cancer screening again highlights that other factors influence CRC screening among Hispanic women. We found that, regardless of eth- nicity, men who were current for prostate cancer screening were also current for CRC screening. Even though Hispanic women were less to be current with CRC screening than White women, we found no ethnic differences in current breast or cervical screening rates.
Beyond SES and access to CRC screening, patient knowledge, attitudes and cultural beliefs about colorectal cancer screening also influence screening. Reported barriers to CRC screening among minority groups include lack of understanding about cancer, screening, and medical terminology as well as concerns about death, and anxiety surrounding the possibility of finding cancer if tested [16]. Ogedegbe et al. [17] noted common barriers to cancer screening among Hispanic and African American women, including lack of cancer screening knowledge, patient’s perception of good health or absence of symptoms, fear of pain from cancer test, and lack of a clinician recommendation. These barriers may be differentially influencing CRC screening versus other cancer screening tests for our population of Hispanic women.
Additional cultural factors, such as including country of origin and level of acculturation (increased use of English), may also impact the prevalence of CRC screening and many CRC-related risk factors. In the current BRFSS analysis, up-to-date CRC screening was less prevalent among Spanish- versus English-speaking Hispanics. Spanish-language preference, in addition to being an indicator for acculturation, is also related to health literacy and cultural beliefs about disease and prevention, which may influence CRC screening uptake [5, 8, 18]. However, the sample size of Spanish-speakers was too small to test for differences in CRC risk factors by language preference.
Data from national surveys also indicate a higher prevalence of physical inactivity, high BMI, smoking and diabetes in Hispanics compared to whites [19]. The National Cancer Institute estimates a 40 % increase in CRC incidence for alcohol consumption exceeding 45 grams/day, an 18 % increased risk associated with cigarette smoking, and a 45 % increase risk in women with a BMI of more than 29 [20]. A 30 % increase in CRC incidence was reported by Larsson et al. [21]. In contrast, regular physical activity was associated with a 45 % reduction in CRC incidence. For New Mexican Hispanics, the potential magnitude of excess CRC risk associated with these factors is substantial.
Differences in socioeconomic measures may also be differentially influencing these lifestyle and clinical measures. Several studies have demonstrated an association between CRC risk factors and income, education, and employment status, measures of which were all lower among Hispanics than among whites in our survey data. For example, these socioeconomic measures have been demonstrated to be positively associated with the intake of fruits and vegetables and physical activity, and a negatively association with cigarette smoking, drinking alcohol, and cancer [22–26]. An association between socioeconomic measures and obesity has also been reported, although the direction of the association differs by gender: men of high socioeconomic status (SES) are more likely to be over- weight than men of low SES, but women of high SES are less likely to be overweight than women of low SES [27]. Because risk factors generally affect disease development over the long term rather than the short term, the current risk factor data potentially reflect years of unhealthy life- styles among Hispanics which has important public health and policy implications for reducing the burden of colorectal cancer in this population.
There are several inherent limitations to drawing inference from BRFSS data. First, the BRFSS is a telephone survey, thus only households with landline telephones are represented, potentially creating a selection bias. Response bias is of concern. Responses are self-reported and not confirmed by medical records review. Because the BRFSS is a cross-sectional survey, we cannot conclude that the ethnic differences we observed in CRC screening and risk factors actually account for the differences in CRC incidence among Hispanics and whites. Second, the design does not allow us to evaluate ethnic differences in adherence with recommended CRC screening tests over time. Last, the BRFSS defines current as having completed a fecal blood test within the past one year and/or lower endoscopy within the past 10 years. However, the appropriate interval for flexible sigmoidoscopy is 5 years, thus the current analysis for New Mexico may overestimate up- to-date CRC screening [3]. A recent survey of primary care physicians in New Mexico, though, suggests that flexible sigmoidoscopy is infrequently offered to patients [27].
Much of the difference in CRC screening observed between Hispanics and white respondents is attributable to socioeconomic factors, which may also confound the association between critical risk factors and CRC. Socioeconomic indicators likely operate through multiple mechanisms that influence the risk of colorectal cancer, and not solely access to health care and screening tests [28]. Health reform will increase access for CRC screening (U.S. Preventive Health Services A rating), which will help overcome the socioeconomic barriers to access [3]. While overcoming these barriers is necessary, it will not be sufficient for achieving an increase in CRC screening. Efforts must also be made to overcome the cultural, linguistic and knowledge barriers, which also influence CRC screening. Results from a recent national survey indicate that health care providers were an extremely important information source for cancer screening [28]. In particular, women were more likely than men to report that health care providers were an extremely important source of CRC screening information. Physicians can also help patients understand the role of risk factors in CRC. Shike et al. [29] found that offering CRC screening to minority women at the time of mammography–without requiring a physician’s referral–was an effective way to increase screening. Innovative approaches are needed to reach Hispanics to raise the awareness of and participation in CRC screening. Because many CRC risk factors are potentially modifiable, cultural and linguistic appropriate interventions tailored to specific the Hispanic subgroups in New Mexico, which promote CRC screening and the reduction of CRC risk factors may decrease ethnic disparities in CRC incidence seen in our state.
Acknowledgments
We wish to thank Wayne Honey, MPH, at the New Mexico Department of Health for providing the New Mexico BRFSS data for this analysis. Dr. Gonzales was supported by the National Institute of Environmental Health Sciences K01 ES014003.
Contributor Information
Melissa Gonzales, Department of Medicine, University of New Mexico School of Medicine, MSC10 5550 Epidemiology, Albuquerque, NM 87131-0001, USA.
Harold Nelson, New Mexico Tumor Registry, University of New Mexico, Albuquerque, NM, USA.
Robert L. Rhyne, Department of Family and Community Medicine, University of New Mexico School of Medicine, Albuquerque, NM, USA
S. Noell Stone, New Mexico Department of Health, Albuquerque, NM, USA.
Richard M. Hoffman, Department of Medicine and Department of Family and Community Medicine, University of New Mexico School of Medicine, Albuquerque, NM, USA; New Mexico Veterans Administration Health Care System, Albuquerque, NM, USA
References
- 1.Edwards BK, Ward E, Kohler BA, Eheman C, et al. Annual report to the nation on the status of cancer, 1975–2006, featuring colorectal cancer trends and impact of interventions to reduce future rates. Cancer. 2010;116(3):544–573. doi: 10.1002/cncr.24760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.New Mexico Cancer Facts and Figures. A document for tracking, planning and evaluating. Albuquerque, NM: New Mexico Department of Health; 2007. http://www.cancernm.org/cancercouncil/pdf/NMCFF_FINAL_8-28-07_reduced%20size.pdf Accessed October 9, 2011. [Google Scholar]
- 3.Whitlock EP, Lin J, Liles E, Beil T, et al. Screening for colorectal cancer: An updated systematic review. Rockville, MD: Agency for Healthcare Research and Quality; 2008. http://www.ncbi.nlm.nih.gov/bookshelf/br.fcgi?book=es65v1 Accessed September 23, 2011. [PubMed] [Google Scholar]
- 4.Cronin CN, Klabunde KN, Breen KA, Waldron N, Ambs AH, Nadel MR. Trends in colorectal cancer test use among vulnerable populations in the United States. Cancer Epidemiology Biomarkers Prevention. 2011;20(8):1611–1621. doi: 10.1158/1055-9965.EPI-11-0220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhou J, Enewold L, Peoples GE, Clifton GT, et al. Trends in cancer screening among Hispanic and white non-His- panic women, 2000–2005. Journal of Womens Health. 2010;19(12):2167–2174. doi: 10.1089/jwh.2009.1909. [DOI] [PubMed] [Google Scholar]
- 6.Zhou J, Enewold L, Peoples GE, McLeod DG, et al. Colorectal, prostate, and skin cancer screening among Hispanic and White non-Hispanic men, 2000–2005. Journal of the National Medical Association. 2011;103(4):343–350. doi: 10.1016/s0027-9684(15)30315-1. [DOI] [PubMed] [Google Scholar]
- 7.Afable-Munsuz A, Liang SY, Ponce NA, Walsh JM. Acculturation and colorectal cancer screening among older Latino adults: Differential associations by national origin. Journal of General Internal Medicine. 2009;24(8):963–970. doi: 10.1007/s11606-009-1022-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA. 2010;303(3):235–241. doi: 10.1001/jama.2009.2014. [DOI] [PubMed] [Google Scholar]
- 9.Centers for Disease Control and Prevention. Behavioral risk factor surveillance system survey data. Atlanta, GA: US Department of Health and Human Services; 2006. [Google Scholar]
- 10.Centers for Disease Control and Prevention (CDC) Use of colorectal cancer tests—United States, 2002, 2004, and 2006. Morbidity and Mortality Weekly Report. 2008;57(10):253–258. [PubMed] [Google Scholar]
- 11.Kang-Kim M, Betancourt JR, Ayanian JZ, Zaslavsky AM, Yucel RM, Weissman JS. Access to care and use of preventive services by Hispanics: State-based variations from 1991 to 2004. Medical Care. 2008;46(5):507–515. doi: 10.1097/MLR.0b013e31816dd966. [DOI] [PubMed] [Google Scholar]
- 12.Peterson NB, Murff HJ, Ness RM, Dittus RS. Colorectal cancer screening among men and women in the United States. Journal of Womens Health. 2007;16(1):57–65. doi: 10.1089/jwh.2006.0131. [DOI] [PubMed] [Google Scholar]
- 13.Vargas Bustamante A, Chen J, Rodriguez HP, Rizzo JA, Ortega AN. Use of preventive care services among Latino subgroups. American Journal of Preventive Medicine. 2010;38(6):610–619. doi: 10.1016/j.amepre.2010.01.029. [DOI] [PubMed] [Google Scholar]
- 14.Swan J, Breen N, Coates RJ, Rimer BK, Lee NC. Progress in cancer screening practices in the United States: Results from the 2000 national health interview survey. Cancer. 2003;97(6):1528–1540. doi: 10.1002/cncr.11208. [DOI] [PubMed] [Google Scholar]
- 15.Shokar NK, Vernon S, Weller SC. Cancer and colorectal cancer: Knowledge, beliefs, and screening preferences of a diverse patient population. Family Medicine. 2005;37(5):341–347. [PubMed] [Google Scholar]
- 16.Ogedegbe G, Cassells AN, Robinson CM, et al. Perceptions of barriers and facilitators of cancer early detection among low-income minority women in community health centers. Journal of the National Medical Association. 2005;97(2):162–170. [PMC free article] [PubMed] [Google Scholar]
- 17.Jerant AF, Arellanes RE, Franks P. Factors associated with Hispanic/non-Hispanic white colorectal cancer screening disparities. Journal of General Internal Medicine. 2008;23(8):1241–1245. doi: 10.1007/s11606-008-0666-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bolen JC, Rhodes L, Powell-Griner EE, Bland SD, Holtzman D. State-specific prevalence of selected health behaviors, by race and ethnicity-behavioral risk factor surveillance system, 1997. Morbidity and Mortality Weekly Report Surveillance Summaries. 2000;49(2):1–60. [PubMed] [Google Scholar]
- 19.Colorectal Cancer Prevention (PDQ®) National Institutes of Health. National Cancer Institute; 2011. http://www.cancer.gov/cancertopics/pdq/prevention/colorectal/HealthProfessional/page2 Updated July 15, 2011. Accessed September 24, 2011. [Google Scholar]
- 20.Larsson SC, Orsini N, Wolk A. Diabetes mellitus and risk of colorectal cancer: a meta-analysis. Journal of the National Cancer Institute. 2005;97(22):1679–1687. doi: 10.1093/jnci/dji375. [DOI] [PubMed] [Google Scholar]
- 21.Barbeau EM, Krieger N, Soobader MJ. Working class matters: socioeconomic disadvantage, race/ethnicity, gender, and smoking in NHIS 2000. American Journal of Public Health. 2004;94(2):269–278. doi: 10.2105/ajph.94.2.269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Clegg L, Reichman ME, Miller BA, et al. Impact of socioeconomic status on cancer incidence and stage at diagnosis: selected findings from the surveillance, epidemiology, and end results: National Longitudinal Mortality Study. Cancer Causes and Control. 2009;20(4):417–435. doi: 10.1007/s10552-008-9256-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Dubowitz T, Heron M, Bird CE, et al. Neighborhood socioeconomic status and fruit and vegetable intake among whites, blacks, and Mexican Americans in the United States. American Journal of Clinical Nutrition. 2008;87(6):1883–1891. doi: 10.1093/ajcn/87.6.1883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Huckle T, You RQ, Casswell S. Socio-economic status predicts drinking patterns but not alcohol-related consequences independently. Addiction. 2010;105(7):1192–1202. doi: 10.1111/j.1360-0443.2010.02931.x. [DOI] [PubMed] [Google Scholar]
- 25.Tucker-Seeley RD, Subramanian SV, Li Y, Sorensen G. Neighborhood safety, socioeconomic status, and physical activity in older adults. American Journal of Preventive Medicine. 2009;37(3):207–213. doi: 10.1016/j.amepre.2009.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhang Q, Wang Y. Socioeconomic inequality of obesity in the United States: do gender, age, and ethnicity matter? Social Science and Medicine. 2004;58(6):1171–1180. doi: 10.1016/s0277-9536(03)00288-0. [DOI] [PubMed] [Google Scholar]
- 27.Hoffman RM, Rhyne RL, Helitzer DL, et al. Barriers to colorectal cancer screening: physician and general population perspectives, New Mexico, 2006. Preventing Chronic Disease. 2011;8(2):A35. [PMC free article] [PubMed] [Google Scholar]
- 28.Hoffman RM, Lewis LC, Pignone MP, et al. Decision-making processes for breast, colorectal, and prostate cancer screening: the DECISIONS survey. Medical Decision Making. 2010;30(5 Suppl):53S–64S. doi: 10.1177/0272989X10378701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shike M, Schattner M, Genao A, et al. Expanding colorectal cancer screening among minority women. Cancer. 2011;117(1):70–76. doi: 10.1002/cncr.25566. [DOI] [PMC free article] [PubMed] [Google Scholar]