During the study period, 1998–2002, the ASR(World) of OPC incidence and mortality were higher in men in PR than among USH, NHB, and NHW men. This higher burden of OPC in PR is consistent with studies performed in the 60's and 70's comparing cancer risk in PR and the USA [12, 25]. Moreover, our study adds the comparison of these rates with two additional racial/ethnic groups (USH and NHB). Contrary to declining incidence and mortality trends observed for men in the other racial/ethnic groups, men in PR had a constant trend for incidence and increasing mortality trends. Although women in PR had lower incidence rates of OPC than NHB and NHW women, they had the highest increase for incidence and the lowest decrease for mortality when compared to the other three racial/ethnic groups. These results highlight a health disparity for PR that warrants further investigation.
The observed high burden of OPC in PR and increasing patterns of disease occurrence in this population are potentially the result of multiple factors, including lifestyle factors, genetics, acculturation, and screening practices within this population. Tobacco smoking explains some of the sex and racial differences in OPC risk . A large case-control study of OPC in the USA showed that variation in the prevalence of use and the risks associated with tobacco and alcohol accounts for much of the higher incidence among men than women and among NHB as compared to NHW. . In PR, it has been estimated that the attributable risk of OPC due to alcohol and tobacco use is around 76% (95%CI = 65–87%) for men and 52% (95%CI: 28–75%) for women . Similar attributable risks have been reported in the USA .
Another study in PR also showed that OPC cancer in PR is strongly associated to liquor consumed straight, suggesting that alcohol concentration is a risk factor for OPC, independent of the total quantity of alcohol consumed . In addition, risks associated with combined exposure to tobacco were also more pronounced when subjects drank liquor straight. In PR, straight liquor intake is reported more commonly among men (41%) than in their NHW counterparts in the USA (18%) . Furthermore, locally produced, homemade rum is commonly consumed in PR and has been suspected to contain potentially carcinogenic contaminants . Nevertheless, a study suggests that the risks associated with drinking homemade rum in this population are similar to those of other types of liquor .
Another factor that may explain the excess of OPC risk is the use of tobacco products. According to the Behavioral Risk factor Surveillance System (BRFSS), 14.5% of the adult population in PR reported cigarette smoking in 1996 and just 12.2% in 2007 . Smaller studies conducted in PR prior to 1996 (1982 and 1989) suggest a declining trend in cigarette consumption. The prevalence of smoking has also consistently declined for the last three decades for NHW, USB, and USH [33, 34]. Even with the declines reported for all the studied groups, PR has had a historically lower prevalence, almost one third, of cigarette smoking than NHW, NHB, and USH, which does not support the higher burden of OPC in our population. Nonetheless, this needs to be interpreted with caution as a delay of several decades is usually apparent between the first time of exposure to tobacco and the development of oral malignances . In addition, given that historical data on tobacco consumption in PR is scarce, its difficult to correlate tobacco use and OPC risk in this population through the use of ecologic measures .
Genetic predispositions in PR should also be considered, as these predispose to the development of OPC at an early age . In a population-based case-control study conducted in PR, the null GSTM1 genotype was associated with a marginally significant decrease in oral cancer risk (est. OR = 0.6, 95%CI: 0.3–1.0) . The same study concluded that oral cancer risk augmented with increasing cigarette use among subjects with the GSTT1-present genotype (est. OR = 9.5, 95%CI: 3.0–30, among the heaviest cigarette users). Thus, further studies are warranted to determine if the frequency of these genetic polymorphisms, among other well known oncogenes, in PR is higher than among USH, NHW, or NHB; or if gene environment interactions vary among these groups.
Some attention also should be drawn to other lifestyle, immunologic, and environmental potential risk factors for OPC in PR, as these could be influencing the increased burden of OPC in this population. Relevant risk factors for OPC include mouthwash use , poor oral hygiene and poor dentition , oral cavity infections and diseases, denture sores, oral mucosal lesions , and sunlight exposure . Also, exposure to sexually transmitted infections, such as the human papilloma viruses (HPV) , herpes viruses , and human immunodeficiency virus (HIV)  could impact OPC risk. Although no population-based data on the prevalence of HPV or herpes viruses has been published for PR that might explain the observed patterns of OPC occurrence, a higher incidence of HPV has been observed in developing countries of Latin America (33.5 per 100,000) and the Caribbean (33.5)  as compared to North America (< 15 per 100,000). Meanwhile, PR ranks fifth among all USA's states and territories in the rate of reported AIDS cases (26.4 per 100,000) . Thus, a potentially high burden of these infections agents could be influencing the burden of OPC in PR, although additional population-based data is warranted.
In addition, dietary insufficiencies, specially fruit and vegetables, and low levels of serum nutrients such as carotenoids  could play a role in the observed results. Several studies have indicated a protective role for OPC of fruits and vegetables, although the specific agent is uncertain. For example, folates (found in a variety of foods, particularly green leafy vegetables, grain products, and orange juice) have long been hypothesized to be related to cancer risk . A study conducted in PR showed that folate intake from fruit decreased OPC risk (p = 0.001) but other dietary folate sources showed no clear association (p > 0.05). In PR, fruit and vegetable consumption (5 or more per day) has decreased in the last decade (1996–2007) from 20.3% to 13.7%; meanwhile, the consumption for the USH, NHW, and NHB population has not showed a decrease and remains higher than that in PR . This low consumption of fruit and vegetables as a direct source of folate may help to explain the higher incidence of OPC in PR. Nonetheless, historic nutritional information for PR is scare, limiting our ability to further explain the observed trends. Future studies should elucidate the role of diet on OPC risk PR.
The annual ASR (World) for mortality of OPC showed a decline in all racial/ethnic groups, except for men in PR. These declines could be explained by the participation in early detection programs, particularly of women [49–51]. This disparity for Puerto Rican men suggests that access to advances in treatment, screening, and early detection are not optimum in this population. In fact, a recent study by Morse found disparities in the detection of very early oral cancer in PR as compared with the USA . Studies in this area are necessary to elucidate this argument. In addition, efforts must be made to increase the prevalence of OPC screening in PR, as early detection of OPC lesions is the most important factor in the prognosis and cure of this cancer.
Due to the acculturation process, it has been suggested that Hispanic migrants have cancer rates similar to NHW [3, 8]. However, our results showed that USH men and women have lower incidence of OPC than NHW but a very similar mortality, a fact that has been explained by several factors, such as differences in health practices and health care access [3, 8]. Among the Hispanic community living in the continental USA, acculturation occurs more markedly when they arrive to this country. Meanwhile, in the Hispanic population living in PR, acculturation has occurred through the close sociopolitical relationship with the USA since 1898 . Given that the population of PR has experienced an acculturation process different from that of USH, the comparison of OPC incidence and mortality between these groups and the NHW and NHB populations is essential for further understanding differences and similarities of disease occurrence between these groups, that permit the identification of health disparities. However, because of the differences in the populations included under the broad heading of "United States Hispanics"  we cannot conclude that acculturation alone can explain the differences observed among Puerto Ricans and USH.