Do the general public get cancer statistics?-a questionnaire survey in Japan.

OBJECTIVES: The public does not always understand key information conveyed by epidemiologists and statisticians. The purpose of this study was to understand the level of public access to, trust in, and comprehension of, cancer statistics through a population-based survey in Japan. METHODS: We used an online research method, requesting online responses to a 15-question questionnaire. The survey was sent to males and females aged 20 years and older, selected by sex, age and prefecture to match the national population proportions shown in the latest census. The final number of valid responses was 10 477. The statistical analyses mainly used χ2 testing. RESULTS: Respondents were not frequently exposed to cancer statistics regardless of sex or age group, nor did they necessarily have confidence in the statistics. The increase of collected information and trust in cancer statistics was aligned with increasing age and cancer exposure. Respondents found Relative Risk and Relative Survival Rate easier to understand and more useful than the Standardized Incidence Ratio. In addition, those with cancer experience, higher income and were elderly gave more accurate responses when asked questions related to cancer incidence and probability of getting cancer. CONCLUSIONS: Our respondents showed limited familiarity with cancer statistical indicators. Enhanced awareness of indicators such as infographics and visual tools has the potential to enhance cancer visibility, thereby promoting public prevention and early detection efforts. Educating cancer patients about pertinent indicators can boost their confidence in managing their condition. Conversely, the introduction of indicators unrelated to the public should be discouraged.

Impact of cancer and other causes of death on mortality of cancer patients: A study based on Japanese population-based registry data.

Cancer registry data provide a very important source of information for improving our understanding of the epidemiology of various cancers. In this work, we estimated the 5-year crude probabilities of death from cancer and from other causes for five common cancers, namely stomach, lung, colon-rectum, prostate and breast, in Japan, using population-based registry data. Based on data on 344 676 patients diagnosed with one of these cancers between 2006 and 2008 in 21 prefectures participating in the Monitoring of Cancer Incidence in Japan (MCIJ) and followed-up for at least 5 years, we used a flexible excess hazard model to compute the crude probabilities of death for different combinations of sex, age and stage at diagnosis. For tumours diagnosed at the distant stage, as well as for regional lung tumours, the vast majority of deaths at 5 years in cancer patients were attributable to the disease itself (although this proportion was only around 60% in older prostate cancer patients). For localised and most regional tumours, the impact of other causes of death on the total mortality increased with age at diagnosis, especially for localised breast, colorectal and gastric cancer. By allowing the partition of the mortality experience of cancer patients into a cancer- and an other-cause-specific component, crude probability of death estimates provide insight into how the impact of cancer on mortality might differ among populations with different background mortality risks. This might be useful for informing discussions between clinicians and patients about treatment options.

The methylation level of a single cancer risk marker gene reflects methylation burden in gastric mucosa.

BACKGROUND: Gastric cancer risk can be accurately predicted by measuring the methylation level of a single marker gene in gastric mucosa. However, the mechanism is still uncertain. We hypothesized that the methylation level measured reflects methylation alterations in the entire genome (methylation burden), induced by Helicobacter pylori (H. pylori) infection, and thus cancer risk. METHODS: Gastric mucosa of 15 healthy volunteers without H. pylori infection (G1), 98 people with atrophic gastritis (G2), and 133 patients with gastric cancer (G3) after H. pylori eradication were collected. Methylation burden of an individual was obtained by microarray analysis as an inverse of the correlation coefficient between the methylation levels of 265,552 genomic regions in the person's gastric mucosa and those in an entirely healthy mucosa. RESULTS: The methylation burden significantly increased in the order of G1 (n = 4), G2 (n = 18), and G3 (n = 19) and was well correlated with the methylation level of a single marker gene (r = 0.91 for miR124a-3). The average methylation levels of nine driver genes tended to increase according to the risk levels (P = 0.08 between G2 vs G3) and was also correlated with the methylation level of a single marker gene (r = 0.94). Analysis of more samples (14 G1, 97 G2, and 131 G3 samples) yielded significant increases of the average methylation levels between risk groups. CONCLUSIONS: The methylation level of a single marker gene reflects the methylation burden, which includes driver gene methylation, and thus accurately predicts cancer risk.

Correcting for heterogeneity and non-comparability bias in multicenter clinical trials with a rescaled random-effect excess hazard model.

In the presence of competing causes of event occurrence (e.g., death), the interest might not only be in the overall survival but also in the so-called net survival, that is, the hypothetical survival that would be observed if the disease under study were the only possible cause of death. Net survival estimation is commonly based on the excess hazard approach in which the hazard rate of individuals is assumed to be the sum of a disease-specific and expected hazard rate, supposed to be correctly approximated by the mortality rates obtained from general population life tables. However, this assumption might not be realistic if the study participants are not comparable with the general population. Also, the hierarchical structure of the data can induces a correlation between the outcomes of individuals coming from the same clusters (e.g., hospital, registry). We proposed an excess hazard model that corrects simultaneously for these two sources of bias, instead of dealing with them independently as before. We assessed the performance of this new model and compared it with three similar models, using extensive simulation study, as well as an application to breast cancer data from a multicenter clinical trial. The new model performed better than the others in terms of bias, root mean square error, and empirical coverage rate. The proposed approach might be useful to account simultaneously for the hierarchical structure of the data and the non-comparability bias in studies such as long-term multicenter clinical trials, when there is interest in the estimation of net survival.

The influence of postscreening follow-up time and participant characteristics on estimates of overdiagnosis from lung cancer screening trials.

We aimed to explore the underlying reasons that estimates of overdiagnosis vary across and within low-dose computed tomography (LDCT) lung cancer screening trials. We conducted a systematic review to identify estimates of overdiagnosis from randomised controlled trials of LDCT screening. We then analysed the association of Ps (the excess incidence of lung cancer as a proportion of screen-detected cases) with postscreening follow-up time using a linear random effects meta-regression model. Separately, we analysed annual Ps estimates from the US National Lung Screening Trial (NLST) and German Lung Cancer Screening Intervention Trial (LUSI) using exponential decay models with asymptotes. We conducted stratified analyses to investigate participant characteristics associated with Ps using the extended follow-up data from NLST. Among 12 overdiagnosis estimates from 8 trials, the postscreening follow-up ranged from 3.8 to 9.3 years, and Ps ranged from -27.0% (ITALUNG, 8.3 years follow-up) to 67.2% (DLCST, 5.0 years follow-up). Across trials, 39.1% of the variation in Ps was explained by postscreening follow-up time. The annual changes in Ps were -3.5% and -3.9% in the NLST and LUSI trials, respectively. Ps was predicted to plateau at 2.2% for NLST and 9.2% for LUSI with hypothetical infinite follow-up. In NLST, Ps increased with age from -14.9% (55-59 years) to 21.7% (70-74 years), and time trends in Ps varied by histological type. The findings suggest that differences in postscreening follow-up time partially explain variation in overdiagnosis estimates across lung cancer screening trials. Estimates of overdiagnosis should be interpreted in the context of postscreening follow-up and population characteristics.

Impact of cumulative body mass index and cardiometabolic diseases on survival among patients with colorectal and breast cancer: a multi-centre cohort study.

BACKGROUND: Body mass index (BMI) and cardiometabolic comorbidities such as cardiovascular disease and type 2 diabetes have been studied as negative prognostic factors in cancer survival, but possible dependencies in the mechanisms underlying these associations remain largely unexplored. We analysed these associations in colorectal and breast cancer patients. METHODS: Based on repeated BMI assessments of cancer-free participants from four European countries in the European Prospective Investigation into Cancer and nutrition (EPIC) study, individual BMI-trajectories reflecting predicted mean BMI between ages 20 to 50 years were estimated using a growth curve model. Participants with incident colorectal or breast cancer after the age of 50 years were included in the survival analysis to study the prognostic effect of mean BMI and cardiometabolic diseases (CMD) prior to cancer. CMD were defined as one or more chronic conditions among stroke, myocardial infarction, and type 2 diabetes. Hazard ratios (HRs) and confidence intervals (CIs) of mean BMI and CMD were derived using multivariable-adjusted Cox proportional hazard regression for mean BMI and CMD separately and both exposures combined, in subgroups of localised and advanced disease. RESULTS: In the total cohort of 159,045 participants, there were 1,045 and 1,620 eligible patients of colorectal and breast cancer. In colorectal cancer patients, a higher BMI (by 1 kg/m2) was associated with a 6% increase in risk of death (95% CI of HR: 1.02-1.10). The HR for CMD was 1.25 (95% CI: 0.97-1.61). The associations for both exposures were stronger in patients with localised colorectal cancer. In breast cancer patients, a higher BMI was associated with a 4% increase in risk of death (95% CI: 1.00-1.08). CMDs were associated with a 46% increase in risk of death (95% CI: 1.01-2.09). The estimates and CIs for BMI remained similar after adjustment for CMD and vice versa. CONCLUSIONS: Our results suggest that cumulative exposure to higher BMI during early to mid-adulthood was associated with poorer survival in patients with breast and colorectal cancer, independent of CMD prior to cancer diagnosis. The association between a CMD diagnosis prior to cancer and survival in patients with breast and colorectal cancer was independent of BMI.

Soy product intake and risk of incident disabling dementia: the JPHC Disabling Dementia Study.

PURPOSE: We evaluated the association between total soy, soy product (natto, miso and tofu) and isoflavone intake and incident disabling dementia in a Japanese population. METHODS: We conducted a population-based prospective study in 18,991 men and 22,456 women. Intake of soy products and isoflavone was calculated using a validated food frequency questionnaire when participants were 45-74 years old (1995 and 1998). Incident disabling dementia was defined by the daily living disability status related to dementia in the long-term care insurance program of Japan from 2006 to 2016. Multivariate hazard ratios (HRs) and 95% confidence intervals (CIs) of disabling dementia were calculated by quintiles of total soy, individual soy product and isoflavone intake, using Cox proportional hazard regression models. RESULTS: Total soy product intake was not associated with disabling dementia risk in both men and women. By individual soy products, natto intake was marginally inversely associated with disabling dementia in women (trend P = 0.050). When we stratified by age, this inverse association was clearer in women aged under 60 years (multivariate HR for the highest versus lowest quintile was 0.78, 95% CI 0.59-1.04, trend P = 0.020 for those aged under 60 years and 0.90, 95% CI 0.77-1.05, trend P = 0.23 for those aged 60 years and older, respectively). Any soy product or isoflavone intake was not associated with disabling dementia risk in men. CONCLUSIONS: Although total soy product intake was not associated with disabling dementia risk, natto intake may contribute to reducing the risk of disabling dementia in women, especially in those aged under 60 years.

Long-term exposure to fine particle matter and all-cause mortality and cause-specific mortality in Japan: the JPHC Study.

BACKGROUND: Many epidemiological studies have reported the association between exposure to particulate matter and mortality, but long-term prospective studies from Asian populations are sparse. Furthermore, associations at low levels of air pollution are not well clarified. Here, we evaluated associations between long-term exposure to particulate matter <2.5 µg/m3 (PM2.5) and mortality in a Japanese cohort with a relatively low exposure level. METHODS: The Japan Public Health Center-based Prospective Study (JPHC Study) is a prospective cohort study of men and women aged 40-69 years in 1990 who were followed up through 2013 for mortality. In this cohort of 87,385 subjects who did not move residence during follow-up, average PM2.5 levels from 1998 to 2013 by linkage with 1-km2 grids of PM2.5 concentration were assigned to the residential addresses of all participants. To avoid exposure misclassification, we additionally evaluated the association between 5-year (1998-2002) cumulative exposure level and mortality during the follow-up period from 2003 to 2013 in 79,078 subjects. Cox proportional hazards models were used to calculate the association of long-term exposure to PM2.5 on mortality, with adjustment for several individual confounding factors. RESULTS: Average PM2.5 was 11.6 µg/m3. Average PM2.5 exposure was not associated with all-cause mortality or cancer and respiratory disease mortality. However, average PM2.5 was positively associated with mortality from cardiovascular disease (hazard ratio (HR) of 1.23 (95%CI=1.08-1.40) per 1-µg/m3 increase; in particular, HR in mortality from cerebrovascular disease was 1.34 (95%CI=1.11-1.61) per 1-µg/m3 increase. Additionally, these results using cumulative 5-year PM2.5 data were similar to those using average PM2.5 over 15 years. CONCLUSIONS: We found evidence for a positive association between PM2.5 exposure and mortality from cardiovascular disease in a Japanese population, even in an area with relatively low-level air pollution.

Global burden of cancer in 2020 attributable to alcohol consumption: a population-based study.

BACKGROUND: Alcohol use is causally linked to multiple cancers. We present global, regional, and national estimates of alcohol-attributable cancer burden in 2020 to inform alcohol policy and cancer control across different settings globally. METHODS: In this population-based study, population attributable fractions (PAFs) calculated using a theoretical minimum-risk exposure of lifetime abstention and 2010 alcohol consumption estimates from the Global Information System on Alcohol and Health (assuming a 10-year latency period between alcohol consumption and cancer diagnosis), combined with corresponding relative risk estimates from systematic literature reviews as part of the WCRF Continuous Update Project, were applied to cancer incidence data from GLOBOCAN 2020 to estimate new cancer cases attributable to alcohol. We also calculated the contribution of moderate (<20 g per day), risky (20-60 g per day), and heavy (>60 g per day) drinking to the total alcohol-attributable cancer burden, as well as the contribution by 10 g per day increment (up to a maximum of 150 g). 95% uncertainty intervals (UIs) were estimated using a Monte Carlo-like approach. FINDINGS: Globally, an estimated 741 300 (95% UI 558 500-951 200), or 4·1% (3·1-5·3), of all new cases of cancer in 2020 were attributable to alcohol consumption. Males accounted for 568 700 (76·7%; 95% UI 422 500-731 100) of total alcohol-attributable cancer cases, and cancers of the oesophagus (189 700 cases [110 900-274 600]), liver (154 700 cases [43 700-281 500]), and breast (98 300 cases [68 200-130 500]) contributed the most cases. PAFs were lowest in northern Africa (0·3% [95% UI 0·1-3·3]) and western Asia (0·7% [0·5-1·2]), and highest in eastern Asia (5·7% [3·6-7·9]) and central and eastern Europe (5·6% [4·6-6·6]). The largest burden of alcohol-attributable cancers was represented by heavy drinking (346 400 [46·7%; 95% UI 227 900-489 400] cases) and risky drinking (291 800 [39·4%; 227 700-333 100] cases), whereas moderate drinking contributed 103 100 (13·9%; 82 600-207 200) cases, and drinking up to 10 g per day contributed 41 300 (35 400-145 800) cases. INTERPRETATION: Our findings highlight the need for effective policy and interventions to increase awareness of cancer risks associated with alcohol use and decrease overall alcohol consumption to prevent the burden of alcohol-attributable cancers. FUNDING: None.

Age disparities in stage-specific colon cancer survival across seven countries: An International Cancer Benchmarking Partnership SURVMARK-2 population-based study.

We sought to understand the role of stage at diagnosis in observed age disparities in colon cancer survival among people aged 50 to 99 years using population-based cancer registry data from seven high-income countries: Australia, Canada, Denmark, Ireland, New Zealand, Norway and the United Kingdom. We used colon cancer incidence data for the period 2010 to 2014. We estimated the 3-year net survival, as well as the 3-year net survival conditional on surviving at least 6 months and 1 year after diagnosis, by country and stage at diagnosis (categorised as localised, regional or distant) using flexible parametric excess hazard regression models. In all countries, increasing age was associated with lower net survival. For example, 3-year net survival (95% confidence interval) was 81% (80-82) for 50 to 64 year olds and 58% (56-60) for 85 to 99 year olds in Australia, and 74% (73-74) and 39% (39-40) in the United Kingdom, respectively. Those with distant stage colon cancer had the largest difference in colon cancer survival between the youngest and the oldest patients. Excess mortality for the oldest patients with localised or regional cancers was observed during the first 6 months after diagnosis. Older patients diagnosed with localised (and in some countries regional) stage colon cancer who survived 6 months after diagnosis experienced the same survival as their younger counterparts. Further studies examining other prognostic clinical factors such as comorbidities and treatment, and socioeconomic factors are warranted to gain further understanding of the age disparities in colon cancer survival.

Prediagnostic circulating inflammation-related biomarkers and gastric cancer: A case-cohort study in Japan.

Gastric cancer is preceded by a chronic inflammatory process. Circulating levels of inflammation-related markers may reveal molecular pathways contributing to cancer development. Our study evaluated risk associations of gastric cancer with a wide range of systemic soluble inflammation and immune-response proteins. We performed a case-cohort analysis within the JPHC Study II, including a subcohort of 410 participants selected randomly within defined age and sex groups, and 414 individuals with incident gastric cancer. Ninety-two biomarkers were measured in baseline plasma using proximity extension assays. Gastric cancer multivariable hazard ratios were calculated for two to four quantiles used as ordinal variables of each biomarker by Cox proportional hazards regression models with age as the time metric. Of 73 evaluable biomarkers, three (CCL11, CCL20 and IL17C) were associated with increased gastric cancer risk and two (CCL23 and MMP1) with reduced cancer risk (Ptrends < 0.05). However, no association was statistically significant after a false discovery rate correction. This study largely expands the range of inflammation molecules evaluated for gastric cancer risk but failed to identify novel associations with this neoplasia.

Genetic and epigenetic alterations in normal tissues have differential impacts on cancer risk among tissues.

Genetic and epigenetic alterations are both involved in carcinogenesis, and their low-level accumulation in normal tissues constitutes cancer risk. However, their relative importance has never been examined, as measurement of low-level mutations has been difficult. Here, we measured low-level accumulations of genetic and epigenetic alterations in normal tissues with low, intermediate, and high cancer risk and analyzed their relative effects on cancer risk in the esophagus and stomach. Accumulation of genetic alterations, estimated as a frequency of rare base substitution mutations, significantly increased according to cancer risk in esophageal mucosae, but not in gastric mucosae. The mutation patterns reflected the exposure to lifestyle risk factors. In contrast, the accumulation of epigenetic alterations, measured as DNA methylation levels of marker genes, significantly increased according to cancer risk in both tissues. Patients with cancer (high-risk individuals) were precisely discriminated from healthy individuals with exposure to risk factors (intermediate-risk individuals) by a combination of alterations in the esophagus (odds ratio, 18.2; 95% confidence interval, 3.69-89.9) and by only epigenetic alterations in the stomach (odds ratio, 7.67; 95% confidence interval, 2.52-23.3). The relative importance of epigenetic alterations upon genetic alterations was 1.04 in the esophagus and 2.31 in the stomach. The differential impacts among tissues will be critically important for effective cancer prevention and precision cancer risk diagnosis.

Cumulative exposure to premenopausal obesity and risk of postmenopausal cancer: A population-based study in Icelandic women.

Obesity, often assessed at one point in time, is an established risk factor of several types of cancer, however, associations with cumulative exposure to obesity across the life course are not well understood. We investigated the relationship between combined measures of duration and intensity of premenopausal overweight and obesity and the incidence of postmenopausal breast, endometrial, and colorectal cancers in Icelandic women. Body mass index (BMI) trajectories between ages 20 and 50 of 88,809 women from the Cancer Detection Clinic Cohort were predicted using growth curve models. Indicators of overweight and obesity duration and intensity were computed and their association with risk of postmenopausal breast, endometrial, and colorectal cancers was examined using multivariate Cox models for subjects followed-up beyond the age of 50 (n = 67,488). During a mean follow-up of 17 years, incident events of 3,016 postmenopausal breast, 410 endometrial and 987 colorectal cancers were ascertained. Each 0.1 kg/m2 per year increase in BMI between ages 20 and 50 was positively associated with risks of postmenopausal breast, endometrium and colorectal cancers with hazard ratios equal to 1.09 (95% Confidence Interval (CI):1.04-1.13), 1.31 (95% CI: 1.18-1.44) and 1.10 (95% CI: 1.00-1.21), respectively. Compared to women who were never obese, cumulative BMI × years of obesity were linearly positively associated with risk of endometrial cancer, whereas the association with breast cancer was initially positive, but leveled off with increasing cumulative BMI × years. Cumulative exposure to obesity may provide additional insights into the etiology of cancer and should be considered in future studies that assess obesity-cancer relationships.

Prediagnostic circulating inflammation biomarkers and esophageal squamous cell carcinoma: A case-cohort study in Japan.

Esophageal squamous cell carcinoma (ESCC) is the predominant histologic subtype of esophageal cancer worldwide. Measurements of circulating inflammation-related biomarkers may inform etiology or provide noninvasive signatures for early diagnosis. We therefore examined levels of inflammation molecules for associations with ESCC risk. Using a case-cohort study designed within the Japan Public Health Center-based Prospective Study, we measured baseline plasma levels of 92 biomarkers using a multiplex assay in a subcohort of 410 randomly selected participants and 66 participants with incident ESCC (including four cases that occurred in the subcohort). ESCC hazard ratios (HRs) were calculated for 2-4 quantiles of each biomarker by Cox proportional hazards regression models with age as the time metric, adjusted for sex, smoking and alcohol use. Twenty analytes were undetectable in nearly all samples. Of the remaining 72, 12 biomarkers (FGF19, ST1A1, STAMBP, AXIN1, CASP8, NT3, CD6, CDCP1, CD5, SLAMF1, OPG and CSF1) were associated with increased ESCC risk (ptrend < 0.05) with HRs per quantile 1.28-1.65. Seven biomarkers (CXCL6, CCL23, CXCL5, TGFA, CXCL1, OSM and CCL4) were inversely associated with HRs 0.57-0.72. FGF19, CASP8, STAMBP, ST1A1 and CCL-4 met statistical significance with false discovery rate correction. Associations did not differ <5 vs. ≥5 years between blood collection and ESCC diagnosis. CASP8, STAMBP and ST1A1 were strongly correlated (p < 0.05). Our study expands the range of inflammation molecules associated with the development of this highly lethal neoplasia. Correlations among these novel biomarkers suggest a possible shared pathway. These findings need replication and could further delineate ESCCs molecular mechanisms of carcinogenesis.

Time trends and other sources of variation in Helicobacter pylori infection in mainland China: A systematic review and meta-analysis.

BACKGROUND: Helicobacter pylori (H pylori) is a carcinogen that causes a huge burden of gastric cancer in China. We aimed to evaluate the temporal trends and other sources of variation of H pylori infection in adults from mainland China. MATERIALS AND METHODS: For this systematic review and meta-analysis, we searched PubMed, Embase, China National Knowledge Infrastructure, and Wanfang databases for articles published from January 1983 to June 2020. We included studies reporting H pylori prevalence in adults and then applied random effect meta-analyses to obtain pooled prevalence estimates for all studies and subgroups. Sources of heterogeneity were investigated by moderator analysis, and time trends were assessed through random effect meta-regression. RESULTS: Of the 2121 studies identified, 98 were eligible for inclusion. The pooled estimate of 670 572 participants from 26 provinces during 1983-2018 was 49.6% (95% CI: 46.9%, 52.4%). H pylori prevalence varied considerably, ranging from 20.6% to 81.8%. Periods, urban/rural status, detection method, and study design explained 18.8%, 24.0%, 17.8%, and 30.4% of the heterogeneity, respectively. Overall, H pylori prevalence declined by -0.9% (95% CI: -1.1%, -0.6%) annually. Consistent declines in prevalence were observed by sex, age, and study characteristics. CONCLUSIONS: Helicobacter pylori prevalence is slowly decreasing over time in mainland China, but the low declining speed is not enough to have a major impact on gastric cancer incidence for many years. The time trends and the large heterogeneity should be taken into account when conducting regional comparisons, disease burden estimations, and customized strategy making.

Cancer incidence in older adults in selected regions of sub-Saharan Africa, 2008-2012.

Although the countries of Sub-Sharan Africa represent among the most rapidly growing and aging populations worldwide, no previous studies have examined the cancer patterns in older adults in the region as a means to inform cancer policies. Using data from Cancer Incidence in Five Continents, we describe recent patterns and trends in incidence rates for the major cancer sites in adults aged ≥60 years and in people aged 0-59 for comparison in four selected population-based cancer registries in Kenya (Nairobi), the Republic of South Africa (Eastern Cape Province), Uganda (Kyadondo country), and Zimbabwe (Harare blacks). Over the period 2008-2012, almost 9,000 new cancer cases were registered in older adults in the four populations, representing one-third of all cancer cases. Prostate and esophageal cancers were the leading cancer sites in older males, while breast, cervical and esophageal cancers were the most common among older females. Among younger people, Kaposi sarcoma and non-Hodgkin lymphoma were common. Over the past 20 years, incidence rates among older adults have increased in both sexes in Uganda and Zimbabwe while rates have stabilized among the younger age group. Among older adults, the largest rate increase was observed for breast cancer (estimated annual percentage change: 5% in each country) in females and for prostate cancer (6-7%) in males. Due to the specific needs of older adults, tailored considerations should be given to geriatric oncology when developing, funding and implementing national and regional cancer programmes.

Plasma C-peptide and glycated albumin and subsequent risk of cancer: From a large prospective case-cohort study in Japan.

To elucidate the individual impacts of insulin and blood glucose on cancer risk, we investigated the association of plasma C-peptide, a surrogated marker of insulin and glycated albumin (GA), a more stable marker of blood glucose, with all-site and site-specific cancer risk by mutually accounting for their confounding effects. The study was prospectively conducted with nearly 4,000 cancer cases arising in our population-based cohort of 33,736 subjects who answered the baseline questionnaire and supplied blood samples. After exclusion of subjects with apparent DM, analysis was done in 3,036 cancer cases and 3,667 subcohort subjects. Among men and women combined, highest levels of C-peptide were statistically significantly associated with an increased risk of all-site [Hazard ratio (HR): 1.21; 95% confidence interval: 1.02-1.42], colon [1.73; 1.20-2.47], liver [3.23; 1.76-5.91], kidney, renal pelvis and ureter cancers [2.47; 1.07-5.69], compared to the respective lowest levels, after adjustment for GA levels. Among these C-peptide-related cancers, colon and liver cancers also showed an increased risk associated with elevated GA levels independently of C-peptide levels. The corresponding HRs for colon and liver cancers compared to the highest and lowest GA levels were 1.43 [1.02-2.00] and 2.02 [1.15-3.55], respectively. Effect modification by gender was only evident for the association between C-peptide and colon cancer (p for interaction = 0.04). Higher insulin levels, independently of higher blood glucose levels, may be relevant to DM-related carcinogenesis for several cancer sites. Examination of circulating insulin levels is a plausible option in evaluating cancer risk even in individuals who have not developed DM.

Estimation of the adjusted cause-specific cumulative probability using flexible regression models for the cause-specific hazards.

In competing risks setting, we account for death according to a specific cause and the quantities of interest are usually the cause-specific hazards (CSHs) and the cause-specific cumulative probabilities. A cause-specific cumulative probability can be obtained with a combination of the CSHs or via the subdistribution hazard. Here, we modeled the CSH with flexible hazard-based regression models using B-splines for the baseline hazard and time-dependent (TD) effects. We derived the variance of the cause-specific cumulative probabilities at the population level using the multivariate delta method and showed how we could easily quantify the impact of a covariate on the cumulative probability scale using covariate-adjusted cause-specific cumulative probabilities and their difference. We conducted a simulation study to evaluate the performance of this approach in its ability to estimate the cumulative probabilities using different functions for the cause-specific log baseline hazard and with or without a TD effect. In the scenario with TD effect, we tested both well-specified and misspecified models. We showed that the flexible regression models perform nearly as well as the nonparametric method, if we allow enough flexibility for the baseline hazards. Moreover, neglecting the TD effect hardly affects the cumulative probabilities estimates of the whole population but impacts them in the various subgroups. We illustrated our approach using data from people diagnosed with monoclonal gammopathy of undetermined significance and provided the R-code to derive those quantities, as an extension of the R-package mexhaz.