When the trends in incidence trends were evaluated by age groups based on life stages (0–19, 20–44, 45–64 and + 65 years), there was a stable and constant increase in incidence in both females and males. Among women, the APCs had little variation. Among men, the highest APC was observed among the youngest group (20 to 44 years). As CRC has a multifactorial etiology, factors such as an increase in weight secondary to changes in dietary patterns and a sedentary lifestyle may contribute to this increase [14]. In countries with a very high HDI, such as the United States, a decrease in the incidence rate of CRC among individuals older than 50 years has been observed, while other groups have increasing or constant APCs of 2.7 (< 40 years) and 1.7 (40 to 49 years) [15]. The decrease in the incidence of CRC in individuals over 50 years is most likely due to public health policies [15]. Because CRC is largely preventable, screenings reduce the incidence among the elderly population; however, as screening is performed earlier, this results in an increase in the incidence rate in the younger population [16].
Regarding mortality, the trends among the groups of men and women over 65 years, between 45 and 64 years and all ages were stable. Studies have shown that in regions with an HDI similar to Brazil, the trends in CRC incidence and mortality are increasing [17]. This reflects the difficulty of access to healthcare, with consequent advanced stages at diagnosis [18, 19], in addition to poor infrastructure, the lack of adequate screening and treatment [20], and the presence of age-related comorbidities [21].
During the study period, women had more incident cases and deaths than men (59.5% and 40.5%; 57.4 and 42.6%, respectively). This finding differs from the results obtained by Ansa et al. [15], who evaluated data from the Surveillance, Epidemiology, and End Results (SEER) Program and observed that CRC was more prevalent among men from 2000 to 2014 in the United States. This difference may be due to the different male: female ratios in the two study areas [22, 23]. In the state of Sergipe, according to the 2010 census, among individuals older than 40 years, the male:female ratio was 1:1.16, while in the United States, the ratio for the same period was 1:1.09 [22, 23].
When the adjusted incidence rates were analyzed, we observed that females and males had similar intermediate levels of variation. These intermediate values occurred due to the epidemiological transition observed in countries with a lower HDI. The highest incidence rates are concentrated in regions with higher HDIs, such as Australia and New Zealand (36.7 cases per 100,000), Europe (28.8–32.1 cases per 100,000), East Asia (26.5 cases per 100,000) and North America (26.2 cases per 100,000) [24]. In regions with lower HDIs, such as Africa and South-Central Asia, the incidence rates are lower (6.4–9.2 per 100,000 and 4.9 per 100,000, respectively) [24].
For both sexes, the left colon and rectum were the most common sites of CRC (Table 1). The epidemiological relevance of these data results from the fact that the most common sites are accessible by flexible rectosigmoidoscopy, which may be a less expensive and more effective screening measure in economically and socially less-developed countries. Ahnen et al. [25] observed a higher prevalence of CRC in the left colon and rectum in the US in 2014, especially among younger individuals.
When the histological subtypes of CRC were considered, approximately 94.5% of the CRCs were adenocarcinoma in both sexes. These values are similar to those described in the international and US literature, in which more than 90% of CRCs are adenocarcinomas [26].
The kernel map showed the geographic distribution, with a hotspot in the northeastern region. This result may be due to the higher population in that region, according to the 2010 census [27]. When analyzing the hotspots according to the 2000 census, there was a small change in the population density [27]. According to the monthly household income distribution by neighborhood based on the 2000 census data, the neighborhoods with the highest monthly household income levels are in the hotspots for the entire study period (1996–2015) [27, 28]. This corroborates the relationship between a higher HDI and a higher incidence of CRC [29]. The neighborhoods with the lowest monthly incomes were also located in the hotspots due to their proximity to the more affluent areas and their population density.
The strengths of our study were as follows: a long study period (1996–2015) and incidence data that were validated nationally and internationally. The database was of high quality, with 94.8% of the diagnoses verified with pathology, and only 3.5% based on death certificates and 1.6% based on C80.