Incidence and mortality rates for colorectal cancer in Puerto Rico and among Hispanics, Non-Hispanic Whites, and Non-Hispanic Blacks in the United States, 1998–2002

Abstract

Background

Colorectal cancer (CRC) is the second most commonly diagnosed cancer in Puerto Rico (PR). In the United States of America (US), incidence and mortality rates of CRC show great variation by sex and race/ethnicity. Age-standardized incidence and mortality rates of CRC in PR and compared them with those of Hispanics (USH), non- Hispanic Whites (NHW) and non-Hispanic Blacks (NHB) in the US for the period 1998- 2002 were assessed. Incidence and mortality trends and relative differences among racial/ethnic groups by sex and age were determined.

Methods

Age-standardized rates [ASR(World)] were based on cancer incidence and mortality data from the PR Central Cancer Registry and SEER using the direct method. Annual percent changes (APC) and relative risks (RR) were calculated using Poisson regression models.

Results

For the 1998–2002 period, the APC of CRC incidence and mortality increased for men in PR, contrary to descending trends observed for other racial/ethnic groups. Overall period rates showed that in both sexes, PR had similar incidence and mortality rates from CRC as USH, but lower than those of NHW and NHB. However, PR men and women aged 40–59 years had higher risk of incidence and mortality when compared to their USH counterparts.

Conclusions

Areas of concern include the increasing trends of CRC in PR and the higher burden of the disease among young Puerto Ricans as compared to USH. Further research should be performed in order to guide the design and implementation of CRC prevention and education programs in PR.

BACKGROUND

Cancer is a major public health problem in the United States of America (US) and other developed countries.¹ In the US, colorectal cancer (CRC) is the third most common cancer in incidence and mortality among men and women; however, both incidence and mortality have declined in recent years.² Although the specific reasons for the observed declines in the US are unknown, it has been debated about whether these can be attributed to improved screening or a change in risk factors among the population.³ In Puerto Rico (PR), CRC is also a common malignancy, being the second most commonly diagnosed cancer among men and women, accounting for 13.3% and 14.0%, respectively, of all cancers diagnosed from 1999–2003.⁴ In terms of mortality, CRC accounted for 11.6% of all male cancer deaths and 13.0% of all female cancer deaths in PR during 2000–2004.⁴

In the US, CRC incidence and death rates vary considerably among racial/ethnic groups, with higher rates observed among non-Hispanic Blacks (NHB) and non-Hispanic Whites (NHW) than among US Hispanics (USH).⁵ Ethnic variations in CRC risk in the US may be explained by differences in cultural and regional characteristics, health-related habits or genetic constitution among these groups.⁶ Other factors that perhaps contribute to ethnic variations in the incidence and mortality rates of CRC include differences in access to screening and access to timely state-of-the-art treatments.¹

Given the limited data available on the burden of CRC in PR and how it compares with other racial/ethnic groups in the US,^7–12 this study assessed age-standardized incidence and mortality rates of CRC in PR and compared them to that of USH, NHW and NHB in the US for the period 1998–2002. In addition, we estimate the relative risk of incidence and mortality of CRC by sex and age groups.

METHODS

Data sources

CRC incidence statistics from 1998 to 2002 for PR were obtained from the Puerto Rico Central Cancer Registry (PRCCR).¹³ The PRCCR is the fourth oldest population-based cancer registry in the world¹⁴ collecting information on cancer in PR since 1951. Data collected includes demographic characteristics, diagnostic information, anatomical site and histology of the tumor, stage at diagnosis, treatment, and date and cause of death of cancer cases. The PRCCR uses the Surveillance, Epidemiology, and End Results (SEER) program and the North American Association of Central Cancer Registries (NAACCR) standards for coding data; thus, the registry is fully comparable with both SEER and NAACCR data. All cancer cases diagnosed since 2001 are reported using the third edition of the International Classification of Disease for Oncology (ICD-O-3).¹⁵ Cases from 1992 to 2000 which were originally reported using previous editions of ICD-O were converted to ICD-O-3 codes. Mortality information for PR from 1998–2002 was obtained from the PRCCR as reported by death certificates prepared by the Puerto Rico Department of Health.¹⁶

CRC incidence statistics from 1998 to 2002 for USH, NHW and NHB were obtained from those released by the SEER program.¹⁷ The SEER program identifies Hispanic ethnicity by a combination of medical record review and matching surnames against a list of Hispanic surnames. The term Hispanic used throughout our report does not account for racial differences within the USH population. CRC mortality information for USH, NHW and NHB from 1998–2002 was obtained from the SEER program as reported by the National Center for Health Statistics (NCHS). US mortality cases were obtained for all states except Connecticut, Maine, Maryland, Minnesota, New Hampshire, New York, North Dakota, Oklahoma and Vermont because of the large number of individuals with unknown origin or ethnicity (≥ 10.0% missing) for several years. Thus, the “Hispanic Index” as developed by the National Cancer Institute (NCI)¹⁸ was used to exclude states where mortality statistics for Hispanics were deemed unreliable.

Statistical Analysis

For each ethnic group, we applied the direct method to compute CRC age-standardized incidence and mortality rates per 100,000 persons, during 1998–2002, using the world standard population. These rates were identified by ASR(World), either for incidence or mortality.¹⁹ To assess the trend of CRC risk from 1998 until 2002, the annual ASR(World) were calculated by sex as follows: $ASR{(\mathit{\text{World}})}_i^k=\sum _{j=1}^{18}{w}_j\frac{d_{ij}^k}{n_{ij}^k}$, where w_j is the proportion of persons in the j-th age group of the standard population (World Population), d^k_ij is the number of cases (new cases or deaths) in the j-th age group for the i-th ethnic group in the k-th year, and n^k_ij is the population in the j-th age group of the i-th ethnic group in the k-th year. The annual percent change (APC) of the ASR(World) was estimated by sex, using the joinpoint regression model.²⁰ The Joinpoint Regression Program from SEER was used for the APC estimation with the following parameters: 1) log transformation of the rate, 2) zero joinpoint model, 3) Poisson model using rate, 4) uncorrelated error model, and 5) Hudson’s method.

To assess racial/ethnic group differences, the ASR(World) were grouped from 1998 until 2002, as follows: $ASR{(\mathit{\text{World}})}_i=\sum _{j=1}^{18}{w}_j\frac{\sum _{k=1998}^{2002}{d}_{ij}^k}{\sum _{k=1998}^{2002}{n}_{ij}^k}$. Then, the ratio of two standardized rates $\left[\raisebox{1ex}{$ASR(\mathit{\text{World}}){\mathit{\text{group}}}_i$}\!\left/ \!\raisebox{-1ex}{$ASR(\mathit{\text{World}}){\mathit{\text{group}}}_j$}\right.\right]$ between two different groups were estimated with their 95% confidence interval,²¹ to assess significant differences in CRC incidence and mortality rates between PR as compared to USH, NHW and NHB. This ratio was denoted as Standardized Rate Ratio (SRR) and we used PR as the reference racial/ethnic group. In addition, relative risks (RR) were estimated with their 95% confidence interval to determine relative differences among the study groups by sex and age (in 5-year age groups) during 1998–2002 using the Poisson regression model.²² The reference racial/ethnic group in the RR estimation was PR. The STATA System release 10.0 (STATA Corp, College Station, TX, USA) was used for the statistical analysis.

RESULTS

Trends of ASR(World)

The annual ASR(World) for incidence and mortality of CRC during the study period (1998–2002) showed different patterns among the racial/ethnic groups studied [Figure 1, A–D]. Men in PR showed an increase in the incidence and mortality rates (APC_incidence= 2.6, APC_mortality= 3.8), in contrast, a decrease was observed among USH, NHW and NHB men. Among women in PR, only the mortality has been reduced (APC_mortality= −2.2); however, the incidence in women is almost constant (APC_incidence= 0.2). The USH population showed the highest reduction in the incidence and mortality trends of CRC among men (APC_incidence= −8.1%; APC_mortality= −6.4%) and women (APC_incidence= −6.5%; APC_mortality= −4.5%).

Figure 1.

Trends for colorectal cancer ASR(world) incidence and mortality rates (per 100,000) for Puerto Rico (PR) and among Non-Hispanic Whites (NHW), Non-Hispanic Blacks (NHB) and US Hispanics (USH), 1998–2002

ASR(World)

The ASR(World) for incidence of CRC (1998–2002) showed that in all racial/ethnic groups, men had higher incidence and mortality rates of CRC than women (p<0.05) [Table 1]. The ratio of ASR(World) showed that NHB and NHW men had a 65% (est. SRR=1.65, 95% CI=1.54, 1.76) and 45% (est. SRR=1.45, 95% CI=1.38, 1.52) higher risk, respectively, of CRC as compared to men in PR. Meanwhile, USH men were 8% more likely than men in PR to be diagnosed with CRC. Among women, no significant differences (p>0.05) in incidence rates were observed between PR and USH, although NHB and NHW females showed a 67% (est. SRR=1.67, 95% CI=1.58, 1.77) and 46% (est. SRR=1.46, 95% CI=1.40, 1.53) higher risk, respectively. In terms of mortality, no significant differences were observed between PR and USH, however NHW and NHB showed a 40% and 83% higher risk of death, respectively, than people in PR.

Table 1.

ASR(World) for incidence and mortality (per 100,000) for colorectal cancer during 1998–2002.

Incidence
Sex	PR	USH	NHW	NHB	SRRa USH vs. PRb	SRRa NHW vs. PRb	SRRa NHB vs. PRb
Male	51.50	55.70	74.47	84.84	1.08(1.01, 1.16)	1.45(1.38, 1.52)	1.65(1.54, 1.76)
Female	38.80	39.61	56.68	64.90	1.02 (0.95, 1.09)	1.46(1.40, 1.53)	1.67(1.58, 1.77)
SRRaMales vs. Femalesb	1.33 (1.25, 1.41)	1.41 (1.30, 1.52)	1.31 (1.29, 1.34)	1.31 (1.23, 1.38)
Mortality
Sex	PR	USH	NHW	NHB	SRRa USH vs. PRb	SRRa NHW vs. PRb	SRRa NHB vs. PRb
Male	24.07	23.95	33.56	44.11	0.99(0.92, 1.08)	1.39(1.30, 1.50)	1.83(1.70, 1.98)
Female	17.49	16.25	24.57	32.03	0.93(0.86, 1.01)	1.40(1.31, 1.51)	1.83(1.70, 1.98)
SRRaMales vs. Femalesb	1.38 (1.24, 1.52)	1.47 (1.39, 1.56)	1.37 (1.35, 1.38)	1.38 (1.34, 1.42)

^aThe ratio of two ASR (World) with 95% confidence interval between parentheses.

^bReference group.

Relative Risks

Sex-specific relative risks for incidence and mortality of CRC in USH, NHW, NHB and PR for the study period 1998–2002 are shown in Figure 2 (A–D). Among men, a higher risk of CRC was observed for NHW aged ≥ 60 years and for USH aged 80+ years when compared to PR [Figure 2, A]. NHB men had higher risk of CRC in all age groups than men in PR. Nonetheless, USH men aged 40–59 years had lower risk of incidence and mortality from CRC than men in PR (p<0.05) [Figure 2, A–B].

Figure 2.

Relative Risks (95% CI) for incidence and mortality of colorectal cancer in Non-Hispanic Whites (NHW), Non-Hispanic Blacks (NHB) and US Hispanics (USH) when compared to Puerto Rico (PR), 1998–2002

Among women, NHW aged ≥ 60 years had a higher risk of CRC than women in PR. USH women aged 40–59 years had a reduced risk (p<0.05) than women in PR [Figure 2, C]. As seen in men, NHB women from all age groups had a higher risk of CRC than their counterparts in PR. Regarding mortality, NHB women from all age groups and NHW women above 50 years had increased risk of death from CRC as compared to women in PR (p<0.05). In contrast, the risk of death for USH women from all age groups was similar to that of women in PR (p>0.05) [Figure 2, D].

DISCUSSION

A measure to assess the cancer burden of minority populations is to compare their incidence rates with those of a reference population.²³ The description of CRC incidence and mortality in the population of PR and its comparison with other racial/ethnic groups in the US is essential for the design of etiologic studies and the identification of strategies for cancer prevention, especially primary prevention in high risk groups. Our study showed age and sex differences in the incidence and mortality from CRC among PR, USH, NHW and NHB racial/ethnic groups. For the 1998–2002 period, the APC of CRC incidence and mortality increased for men in PR, contrary to descending trends observed for other racial/ethnic groups. People in PR had similar incidence and mortality rates from CRC as USH, but lower than those of NHW and NHB. However, men and women in PR aged 40–59 years had higher incidence and mortality of CRC than their USH counterparts (p<0.05). The lower incidence and mortality of CRC in PR as compared to NHW observed in our study are consistent with studies performed in the 60’s and 70’s comparing cancer risk in PR and the US;^7–8 these observations support that PR continues to maintain lower levels of CRC incidence and mortality as compared to NHW. Moreover, our study adds the comparisons of these rates with two additional racial/ethnic groups (USH and NHB).

Regardless of the observed lower incidence and mortality rates for CRC among Puerto Ricans when compared with NHW and NHB, the trends for incidence and mortality among the Puerto Rican population show an increase for the period 1998–2002. These trends in PR are consistent with historical data from the PRCCR, which show annual increases in CRC incidence (1987–2003) and mortality (1987–2004) among men and women in PR.⁴ Of special concern is the fact that in contrast to trends in PR, incidence and mortality rates decreased for NHW, NHB and USH during the same time period (1998–2002), and highlights a health disparity that warrants further investigation.

The observed increase in the incidence trends of CRC in PR is potentially the result of multiple factors, including acculturation, lifestyle factors, environmental exposures, genetics and screening practices within this population. Similar increases have been observed in PR for the incidence rates of other cancer types²⁴ and support the notion that as our population acquires western lifestyles, cancer risk is likely to follow those of industrialized societies.²⁵ In addition, the higher incidence of CRC observed among the younger cohort in PR (persons aged 40–59 years) than among USH could imply that young Puerto Ricans may be more acculturated than younger Hispanics in the US, who tend to be Latin American immigrants. Epidemiologic studies show that people who migrate from one country to another quickly adopt the lifestyle and chronic disease patterns of their new host country.^26–28 Despite the fact that Puerto Ricans living in PR do not live physically in the continental US, they have gradually experienced an acculturation process due to their political, economical and social interrelationships with the US that began in 1898.⁷ However, because of the differences in the populations included under the broad heading of “United States Hispanics”²⁹, we can not conclude that acculturation alone can explain the differences observed among young Puerto Ricans and USH. According to the US Census, in 2001, 66% of Hispanics in the US were Mexicans; followed by Puerto Ricans (9%), Cubans (4%) and other ethnic subgroups (21%).²⁹

Although analytic epidemiologic studies of factors related to CRC risk in PR have not been performed, cross-sectional studies provide information on the prevalence of lifestyle factors that might partially explain the observed trends of CRC within our population. Protective factors among the Puerto Rican population, such as lower smoking and alcohol consumption, may explain the lower incidence rates of CRC in PR than among NHB and NHW.^30–31 However, increases in the prevalence of CRC risk factors in PR in the last decades, such as diabetes, obesity, physical inactivity and low consumption of fruits and vegetables could explain the observed increases in incidence for the study period.^30–33 In addition, there is very limited published data regarding genetic or epigenetic factors in PR that might explain the observed patterns of CRC,³⁴ such as the prevalence of mutations in the DNA mismatch repair genes (MLH1, MSH2 or MSH6) associated with Hereditary Non-Polyposis Colorectal Cancer (HNPCC) syndrome.^35–36 These studies are of particular relevance in PR, as the Puerto Rican population (similar to other Hispanic subgroups)³⁷ resulted from the admixing of the genomes of Spaniards (Europeans), Africans and Taínos (Native Americans).³⁸ Studies are currently underway to examine gene-gene and gene-environmental interactions among young CRC patients in PR, including evaluation of familial CRC kindreds.

Disparities in CRC screening among the studied racial/ethnic groups may also be a determining factor in the observed differences in CRC trends observed in our study, as regular CRC screening has been shown to reduce CRC incidence and mortality.³⁹ The U.S. Preventive Task Force and other national organizations recommend that persons aged ≥ 50 years at average risk be screened for CRC using one or more of the following methods: annual fecal occult blood testing (FOBT), sigmoidoscopy or double-contrast barium enema every 5 years, or colonoscopy every 10 years.^40–41 Despite these recommendations, the use of CRC screening procedures remains low, especially among Hispanics. In PR, adherence rates to CRC screening recommendations are also low (37.8%) and below those of USH (46.6%), NHW (59.1%) and NHB (59.1%).^30–31,42 Low adherence rates to colorectal cancer screening may result in late stage at diagnosis, higher mortality and thus poorer disease outcomes. Thus, efforts must be made to increase the prevalence of screening for CRC in all racial/ethnic groups, particularly among Puerto Rican men who exhibited increasing incidence and mortality trends. CRC prevention and education in PR are necessary at all levels, from the health care providers to the general population, in order to effect change.

Despite the lower burden of disease in PR when compared to US, CRC continues to be a public health priority in the Puerto Rican population, being the second leading cancer type in Puerto Rican men and women (after prostate and breast cancer).⁴³ To our knowledge, this is the first epidemiological study that determines relative differences of CRC in Puerto Ricans living in PR and compares them with USH, NHW and NHB in the US using the Cancer Registry Data from PR and SEER. Analytical epidemiologic studies are necessary to understand the higher burden of CRC among younger cohorts in PR as well as the increase in the total incidence of CRC in this population as compared to the marked decrease observed among other racial/ethnic groups in the US (USH, NHW, and NHB). These studies should explore risk factors for CRC occurrence among Puerto Ricans, including the influence of genetics and acculturation.