Cancer is one of the major non-communicable diseases and an important public health focus. According to GLOBOCAN, in 2018, China diagnosed about 4.28 million new cancer cases and 2.86 million cancer deaths, accounting for nearly 24% of all cancer cases and 30% of cancer deaths worldwide at that time.
As early as the 1960s, some population-based studies found that people with a history of type 2 diabetes had a higher risk of developing cancer. In 2009, Vigneri et al. reported an increased risk of liver, pancreatic, endometrial, colon/rectal, and breast cancers in patients with diabetes. Its pathogenesis may be related to hyperglycemia, hyperinsulinemia, insulin resistance and obesity. Recent studies on the relationship between hyperglycemia, hyperinsulinemia/insulin resistance and cancer have not only focused on diabetic patients, but also expanded to non-diabetic participants with fasting hyperglycemia and hyperglycated hemoglobin (HbA1c).
However, the results of these studies relate only to measures related to basal blood glucose and insulin. Due to the complexity of requiring multiple blood samples, the area of insulin under the curve of the oral glucose tolerance test (OGTT) (insulin-AUC) has rarely been studied.
Insulin resistance and whether and how blood insulin and glucose contribute to cancer development during progression from non-diabetic mild hyperglycemia to diabetes have not been well studied. In particular, no studies using the initial cohort have been conducted in China. It is unclear whether the carcinogenic effects in people with diabetes are caused by insulin resistance, hyperinsulinemia, or hyperglycemia itself.
This study aimed to explore predictors of cancer in Chinese adults diagnosed with diabetes or prediabetes or normal glucose tolerance (NGT) in 1986 and followed up for up to 30 years.
This study screened 1700 participants with different glycemic status from 110,660 residents in Daqing, China since 1985, followed up 30 years for cancer results. According to the 1985 WHO diagnostic criteria for type 2 diabetes and IGT, 630 people were identified as newly diagnosed type 2 diabetes mellitus (NDM) and 576 were identified as IGT. An additional 520 NGT subjects, age- and sex-matched to the IGT group, were included in the control group. All participants in this study had no history of cancer at enrollment.
Except for plasma glucose levels, there were significant differences in almost all baseline parameters including age, gender, BMI, blood pressure (SBP) and smoking status among the three groups (NDM, IGT, NGT). Fasting and post-glucose-load plasma insulin responses were also different between the IGT and NGT groups (Table 1). During the 30-year follow-up period (1986-2016), the cumulative incidence of diabetes was 76.9% in the IGT intervention group, 30.4% in the NGT group, and 91.3% in the IGT non-intervention group.
1. Association of diabetes status and blood glucose levels with cancer incidence in the entire study population and in the non-diabetic group< /p>
A total of 259 (15.2%) cancer cases were identified in the three groups during a mean follow-up period of 27 years (IQR: 15-30 years). Among them, 131 cases (50.6%) of gastrointestinal and hepatobiliary cancer (including 74 cases of gastrointestinal cancer, 57 cases of liver, gallbladder, and pancreatic cancer), 66 cases (25.5%) of lung cancer, and 35 cases (13.5%) of genitourinary system cancer ) (including 20 cases of gynecological tumors); and 27 cases (10.4%) of other types of cancer, including bone cancer, skin cancer, brain cancer, oral cancer and blood cancer. The incidence of cancer in the normal glucose tolerance (NGT), impaired glucose tolerance (IGT), and diabetes groups was 6.06, 6.77, and 7.18 per 1000 person-years, respectively (P = 0.02) (P = 0.02) (Table 1).
Across study participants, the results of Cox model analysis showed that after adjusting for age, sex, BMI, SBP and smoking status, plasma glucose levels at 1 and 2 hours during the 75g glucose load OGTT were significantly different from was significantly associated with increased cancer incidence (aHR = 1.03, 95% CI 1.01-1.06, P = 0.02 for 1-hour glucose, and aHR = 1.03, 95% CI 1.004-1.06, P = 0.02 for 2-hour glucose), while Fasting plasma glucose was not significantly associated (aHR = 1.01, 95%CI 0.96-1.07, P = 0.69).
For fluctuations in blood glucose levels, a 1 mmol increase in plasma glucose levels at 1 and 2 hours during OGTT was also associated with a higher risk of cancer (P < 0.003), while in non-GGT including NGT and IGT In diabetic subjects, only a 2-hour increase in glucose levels was associated with cancer risk (P = 0.0049), but at 30 years of follow-up, time without diabetes was inversely associated with cancer risk (10 years increase, aHR = 0.75 , 95%CI 0.64-0.87, P = 0.0002) (Table 2).
Population-wide multivariate Cox regression analysis also revealed that age at baseline (aHR = 1.07, 95% CI 1.05–1.08, P < 0.0001), male ( aHR = 1.36, 95% CI 1.02–1.81, P = 0.04) and smoking status (aHR = 1.44, 95% CI 1.10–1.90, P = 0.009) were significantly associated with increased cancer incidence. After adjusting for age, sex, BMI, and smoking status, cancer risk in the diabetes group was significantly associated with cancer occurrence (aHR = 1.49, 95% CI 1.08-2.04, P = 0.015) (Figure 1).
However, after further adjustment for SBP, the risk of cancer in the diabetes group was only slightly higher than that in the NGT group (aHR = 1.36, 95%CI 0.99-1.88, P = 0.06) (Table 3), although the diabetes group cancer-free time was 7.0 years less in the NGT group (19.2 years vs. 26.2 years, P<0.0001) (Table 1).
Competing risk of death is an issue to consider in cancer risk studies. Some of the results of this study have drawn attention to this issue. Over 30 years of follow-up, all-cause and cardiovascular deaths were observed to more than triple in the NGT, IGT, and NDM groups (29.6% to 45.3% to 68.3% for all deaths and 11.9% to 68.3% for CVD deaths, respectively). 22.4% to 34.2%), while the incidence of deaths decreased gradually among the three groups (36.8% to 27.2% to 18.0%). In addition, the median age of cancer onset in the NGT group (68 years, IQR 57-73 years) overlapped with the age of death in the IGT group (68 years, IQR 62-76 years) and even higher than that in the NDM group (67 years, IQR 59- 71) (Supplementary Table 1). These results highly suggest that in patients with IGT and diabetes, increased premature death from hyperglycemia may mask the true risk of cancer.
As expected, in the Fine-Gray model with all-cause death as a competing risk, the IGT group (aHR = 1.77, 95%CI 1.38-2.27, P < 0.0001) and the diabetes group ( aHR = 3.34, 95% CI 2.64-4.22, P<0.0001) cancer risk was significantly higher compared with the NGT group (Table 3).
2. Association of insulin resistance, plasma insulin levels and diabetes progression with cancer risk in IGT patients
< p> In the IGT group, log-transformed insulin AUC levels during OGTT were significantly associated with cancer risk (aHR = 2.03, 95% CI 1.08-3.81, P = 0.03) (Table 2). In addition, after controlling for age, gender, smoking status, BMI, SBP and lifestyle intervention, multivariate Cox analysis showed that in IGT subjects, progression to diabetes (aHR = 2.28, 95%CI 1.24-4.20, P = 0.008) and insulin area under the curve at baseline (aHR = 1.39 for 1 SD increase, P = 0.02) were also associated with covariate-adjusted cancer risk.
3. Effects of hypoglycemic therapy on cancer risk
and in patients not using these drugs In contrast, the use of insulin and OHA or either drug alone was not associated with an increased risk of developing cancer (aHR = 1.04, 95% CI 0.65-1.68, P = 0.87). Likewise, patients on insulin or insulin + OHA did not have a higher risk of cancer compared with patients on OHA alone (aHR = 0.88, 95% CI 0.50-1.57, P = 0.67) (Supplementary Table 2).
In summary, cancer was found in 15.2% (259/1700) of the participants. The incidence of cancer in the normal glucose tolerance (NGT), impaired glucose tolerance (IGT), and diabetes groups was 6.06, 6.77, and 7.18 per 1000 person-years, respectively (P = 0.02). In the Fine-Gray model with all-cause death as a competing risk, cancer risk was significantly higher in the IGT and diabetes groups compared with the NGT control group (adjusted hazard ratio (aHR) = 1.77, 95% for the IGT group CI 1.38-2.27, P < 0.0001; for diabetes, aHR = 3.34, 95% CI 2.64-4.22, P < 0.0001). In IGT subjects, progression to diabetes (aHR = 2.28, 95% CI 1.24-4.20, P = 0.008) and insulin area under the baseline curve (aHR = 1.39 for 1 SD increase, P = 0.02) were also associated with synergy variable-adjusted cancer risk.
In this population-based prospective cohort of Chinese adults, participants with diabetes or prediabetes in 1985 were associated with a significantly increased risk of cancer at up to 30 years of follow-up. In conclusion, people with diabetes and prediabetes have an increased risk of cancer. In IGT participants, insulin resistance, compensatory hyperinsulinemia, and progression of diabetes accelerated cancer development. It suggests that appropriate hypoglycemic therapy in people with overt diabetes, as well as preventing or delaying the onset of diabetes by eliminating high insulin requirements in prediabetic populations, may help reduce cancer risk.
References
He S, et al. Cancer and its predictors in Chinese adults with newly diagnosed diabetes and impaired glucose tolerance (IGT): a 30-year follow-up of the Da Qing IGT and Diabetes Study. Br J Cancer. 2022 Mar 7. doi: 10.1038/s41416-022-01758-x. Epub ahead of print. PMID: 35256755.