Cancer is currently one of the major diseases that threaten the health of residents (1), and 8–18% cancer patients have diabetes as a comorbid medical condition (2). According to the epidemiological survey conducted by the Chinese Medical Association, the total number of people with diabetes in mainland China is 129.8 million (3). China has become the country with the highest incidence of diabetes in the world (4). Diabetes were significantly related with higher cancer occurrence and mortality in many cancer types (5). Overall cancer risk was found significantly elevated with a standardized increased ratio of 1.15 (95% CI 1.12–1.19) and 1.25 (95% CI 1.21–1.30) in males and females, respectively (6). A systemic review of 23 studies demonstrated a 41% increased risk for long-term, all-cause mortality for diabetic patients compared with those without diabetes (7). Diabetes and associated metabolic disorders contribute directly or indirectly to cancer progression (8). Anaerobic glycolysis, known as Warburg, is classic in cancer. But Warburg does not mean glucose is unimportant. In contrast, hyperglycemia stimulates cancer proliferation (9). Most cancers predominantly express the glucose transporter 1, which has a high affinity for glucose. The increased glycolysis in cancer cells provides the materials necessary for nucleotide, amino acid, and lipid synthesis. Besides, advanced glycation end products (AGRs) due to hyperglycemia, and its receptors (RAGRs) have been reported to lead to oxidative stress and increased inflammation, which promotes cancer growth, angiogenesis, and metastases (10). Obesity, also a prevalence worldwide, is a risk factor of both diabetes and cancer (11, 12). Excessive adipose tissue generates oxidative stress by increased production of pro-inflammatory adipokines. And BMI is regarded as a typical measurement of obese. There have been numerous studies on the relationship between diabetes and cancer. However, at present, barely convinced evidence clearly explains the extent of the effect of diabetes in different cancer types, stages and body mass indices (BMI), which remains to be further investigated (13–15). In addition, as a wasting disease, the nutritional status of patients gradually deteriorates with the development of cancers (16). Therefore, this large-scale retrospective cohort study is to investigate the impact of diabetes on the prognosis of tumor patients.

Among the 2,929 cancer patients recruited, 43.4% were men and 56.6% were women, with a mean age of 55 years. The mean follow-up period was 36.24 months and 791 patients died. According to the ADA criteria for the management of DM, 2,533 patients were included in the non-DM group and 396 patients were included in the DM group. The demographic, clinical and pathological characteristics of patients in the non-DM and DM groups were shown in Table 1. DM was significantly associated with age, PG-SGA, VFA, tumor types and stages (P < 0.05).

In previous studies, significant differences in inflammatory status, nutritional status, and quality of life between diabetic and non-diabetic patients have been demonstrated (17, 18). But there was no large-scale data on the differences between diabetic and non-diabetic patients and whether effective glycemic control can make up the adverse effects of diabetes in cancer patients. In the present study, we found that body composition and inflammatory parameters in cancer patients with diabetes differed from those without diabetes, suggesting that diabetes exacerbated the systemic inflammatory response of the body (19). For diabetic patients with good glycemic control, there still existed relatively more active inflammatory status compared to patients without diabetes, especially NLR. But quite a few indicators, such as albumin, PAB and CRP were not statistically different, suggesting that good glycemic control can reduce the adverse effects of diabetes to some extent.

DM has a negative impact on tumor patients in different stages. For most types of tumors, the prognosis of patients with DM is poor (20). Diabetes has a greater impact on patient prognosis in early stage liver cancer patients due to the combination of systemic metabolic changes and the development of systemic inflammation in patients with liver cancer at early stages (21). However, DM behaved as a stronger risk factor in advanced patients compared with those in early stage in colon cancer, lung cancer and breast cancer, which may due to the longer survival period of advanced cancer that allows the risk of DM to unfold. Therefore, the management of blood glucose in cancer patients with relatively longer survivals becomes increasingly important (22). It is worth mentioning that the adverse effects of diabetes were observed in leukemia, which may further illustrate the role of abnormal metabolism in malignant hematologic diseases and provide a theoretical basis for subsequent studies on the mechanisms of hematologic metabolism.

The BMI stratification was also discussed, and the risk of diabetes was more pronounced in patients with normal BMI. Patients with normal BMI and diabetes tend to have a longer disease duration and are less tolerant of treatment, while patients with higher BMI even without diabetes continuously exist a similar systemic inflammatory response as diabetic patients (23), which may account for the differences in diabetes risk across populations with different BMIs.

There is a limitation for choosing blood glucose as a marker of glycemic control mainly for the shortcomings of possibly inaccurate assessment of long term blood glucose control. The reasons are listed following. First, the glycated hemoglobin A1C (HbA1c), a reliable measurement of glycemic control, was not chose in this study as a marker of glycemic control mainly due to the inconvenience. In real clinical settings, HbA1c was mainly tested when DM was newly diagnosed or a DM patient with unsatisfied blood glucose levels. Second, although the cutoff value of HbA1c for DM is clearly defined, the reasonable threshold of HbA1c for predicting the prognosis was not a consensus (24), which may bring bias if HbA1c was chose as a glycemic control marker. Third, measurements of DM which also have an impact on cancer development should be taken into analysis as confounders. Several diabetes medications are reported related with cancer prognosis, typically metformin. To be accurate, the mechanisms are still unclear and the results of metformin in cancer prognosis are not always positive (25, 26). Given the limited patients using metformin (only 1.2%) in this study, the stratified sub-analysis was not committed. This topic should be further investigated.

In summary, we found that patients with diabetes tend to combine worse body composition and inflammatory indicators, and that glycemic control can ameliorate the impairment of diabetes to some extent. The above results suggest the negative influence of hyperglycemia in systemic inflammation metabolism. Besides, the risk posed by diabetes is not the same in patients with different tumor types and stages. Thus, the management of diabetes should be emphasized, especially for patients in early stages, which may bring a more durable disease-free state. Finally, we analyzed patients with different BMI to further analyze the relationship between body composition and diseases, and we believe that more studies should be done in the future.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

XL, KZ, and WJ conducted and drafted the manuscript. WZ and YL collected and analyzed the data. ML, JC, and WL designed the manuscript. ML, JC, WL, XL, KZ, and WJ revised the manuscript. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

This research was supported by Jilin Province Health Technology Innovation Project (No. 2017J064).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.