Advances in statistical methods for cancer surveillance research: an age-period-cohort perspective

Background

Analysis of Lexis diagrams (population-based cancer incidence and mortality rates indexed by age group and calendar period) requires specialized statistical methods. However, existing methods have limitations that can now be overcome using new approaches.

Methods

We assembled a "toolbox" of novel methods to identify trends and patterns by age group, calendar period, and birth cohort. We evaluated operating characteristics across 152 cancer incidence Lexis diagrams compiled from United States (US) Surveillance, Epidemiology and End Results Program data for 21 leading cancers in men and women in four race and ethnicity groups (the "cancer incidence panel").

Results

Nonparametric singular values adaptive kernel filtration (SIFT) decreased the estimated root mean squared error by 90% across the cancer incidence panel. A novel method for semi-parametric age-period-cohort analysis (SAGE) provided optimally smoothed estimates of age-period-cohort (APC) estimable functions and stabilized estimates of lack-of-fit (LOF). SAGE identified statistically significant birth cohort effects across the entire cancer panel; LOF had little impact. As illustrated for colon cancer, newly developed methods for comparative age-period-cohort analysis can elucidate cancer heterogeneity that would otherwise be difficult or impossible to discern using standard methods.

Conclusions

Cancer surveillance researchers can now identify fine-scale temporal signals with unprecedented accuracy and elucidate cancer heterogeneity with unprecedented specificity. Birth cohort effects are ubiquitous modulators of cancer incidence in the US. The novel methods described here can advance cancer surveillance research.

Trends and projections of cause-specific premature mortality in Australia to 2044: a statistical modelling study

Background

Long-term projections of premature mortality (defined as deaths age <75 years) help to inform decisions about public health priorities. This study aimed to project premature mortality rates in Australia to 2044, and to estimate numbers of deaths and potential years of life lost (PYLL) due to premature mortality overall and for 59 causes.

Methods

We examined the past trends in premature mortality rates using Australian mortality data by sex, 5-year age group and 5-year calendar period up to 2019. Cigarette smoking exposure data (1945-2019) were included to project lung cancer mortality. Age-period-cohort or generalised linear models were developed and validated for each cause to project premature mortality rates to 2044.

Findings

Over the 25-year period from 1990-1994 to 2015-2019, there was a 44.4% decrease in the overall age-standardised premature mortality rate. This decline is expected to continue, from 162.4 deaths/100,000 population in 2015-2019 to 141.7/100,000 in 2040-2044 (12.7% decrease). Despite declining rates, total numbers of premature deaths are projected to increase by 22.8%, rising from 272,815 deaths in 2015-2019 to 334,894 deaths in 2040-2044. This is expected to result in 1.58 million premature deaths over the 25-year period 2020-2044, accounting for 24.5 million PYLL. Of the high-level cause categories, cancer is projected to remain the most common cause of premature death in Australia by 2044, followed by cardiovascular disease, external causes (including injury, poisoning, and suicide), and respiratory diseases.

Interpretation

Despite continuously declining overall premature mortality rates, the total number of premature deaths in Australia is projected to remain substantial, and cancer will continue to be the leading cause. These projections can inform the targeting of public health efforts and can serve as benchmarks against which to measure the impact of future interventions. They emphasise the ongoing importance of accelerating the prevention, early detection, and treatment of key health conditions.

Funding

No funding was provided for this study.

Trends and Projection of the Incidence of Active Pulmonary Tuberculosis in Southwestern China: Age-Period-Cohort Analysis

BACKGROUND

The control of pulmonary tuberculosis (PTB) is critical for achieving the vision of World Health Organization's End TB goal.

OBJECTIVE

This study analyzes the temporal trends in PTB incidence associated with age, period, and birth cohorts from 2006 to 2020 in Yunnan, China; projects the PTB burden till 2030; and explores the drivers of PTB incidence.

METHODS

The aggregated PTB incidence rates between 2005 and 2020 were obtained from the National Notifiable Disease Reporting System. We used the age-period-cohort model to evaluate the age, period, and cohort effects on PTB incidence. We applied the Bayesian age-period-cohort model to project future PTB incidence from 2021 to 2030. We applied the decomposition algorithm to attribute the incidence trends to population aging, population growth, and age-specific changes from 2006 to 2030.

RESULTS

From 2006 to 2020, the PTB incidence in Yunnan was relatively stable, although the absolute number showed an increase. The net drift was -1.56% (95% CI -2.41% to -0.70%). An M-shaped bimodal local drift and a longitudinal age curve were observed. The overall local drift was below zero for most age groups except for the age groups of 15-19 years (2.37%, 95% CI -0.28% to 5.09%) and 50-54 years (0.41%, 95% CI -1.78% to 2.64%). The highest risk of PTB incidence was observed in the age group of 65-69 years, and another peak was observed in the age group of 20-24 years. Downward trends were observed for both period and cohort effects, but the cohort effect trends were uneven. A higher risk was observed for the birth cohorts of 1961-1970 (rate ratio [RR]1961-1965=1.10, 95% CI 0.88-1.38; RR1966-1970=1.11, 95% CI 0.92-1.37) and 2001-2010 (RR2001-2005=0.92, 95% CI 0.63-1.34; RR2006-2010=0.84, 95% CI 0.45-1.58) than for the adjacent cohorts. The Bayesian age-period-cohort model projected that PTB incidence will continually increase from 2021 to 2030 and that PTB incidence in 2030 will be 2.28 times higher than that in 2006. The age-specific change was the leading cause for the growing PTB disease burden.

CONCLUSIONS

Although there are several levels and measures for PTB control, the disease burden is likely to increase in the future. To bridge the gap of TB-free vision, our study suggests that public health policies be put in place soon, including large-scale active case-finding, priority prevention policies for high-risk older adult and young adult populations, and reduction of possible grandparent-grandchildren transmission patterns.

Secular trends of morbidity and mortality of thyroid cancer in five Asian countries from 1990 to 2019 and their predictions to 2035

BACKGROUND

The rising burden of thyroid cancer (TC) is a public health problem in Asia. Predicting the future burden of TC in Asian countries is essential for disease prevention.

METHODS

Data were obtained from the Global Burden of Disease 2019 for five Asian countries. We applied Bayesian age-period-cohort models to predict morbidity and mortality to 2035 and calculated age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR). Furthermore, the estimated annual percentage change was calculated to evaluate the variation of ASIR and ASMR.

RESULTS

By 2035, predictions suggest that cases of TC will reach 75.56 × 103 in China, 70.22 × 103 in India, 15.78 × 103 in the Republic of Korea, 9.01 × 103 in Japan and 5.55 × 103 in Thailand, respectively. Except Japan, a significant upward trend of ASIR of TC will be observed in five Asian countries. The deaths from TC will increase in five countries and India will become the highest reaching 14.07 × 103 . The ASMR will rise to 0.83/100 000 in India and 1.06/100 000 in the Republic of Korea, while it will drop to 0.35/100 000 in China, 0.43/100 000 in Japan and 0.50/100 000 in Thailand. In further predictions projected by sex, the growth rate of ASIR is reported higher in males than in females among most countries. ASMR of male will exceed that of females in China and Thailand by 2035.

CONCLUSION

The disease burden caused by TC will further increase in five Asian countries, especially for men. It is necessary to develop more rational and timely disease prevention and manage strategies facing this disease trend.

Comparative age-period-cohort analysis

BACKGROUND

Cancer surveillance researchers analyze incidence or mortality rates jointly indexed by age group and calendar period using age-period-cohort models. Many studies consider age- and period-specific rates in two or more strata defined by sex, race/ethnicity, etc. A comprehensive characterization of trends and patterns within each stratum can be obtained using age-period-cohort (APC) estimable functions (EF). However, currently available approaches for joint analysis and synthesis of EF are limited.

METHODS

We develop a new method called Comparative Age-Period-Cohort Analysis to quantify similarities and differences of EF across strata. Comparative Analysis identifies whether the stratum-specific hazard rates are proportional by age, period, or cohort.

RESULTS

Proportionality imposes natural constraints on the EF that can be exploited to gain efficiency and simplify the interpretation of the data. Comparative Analysis can also identify differences or diversity in proportional relationships between subsets of strata ("pattern heterogeneity"). We present three examples using cancer incidence from the United States Surveillance, Epidemiology, and End Results Program: non-malignant meningioma by sex; multiple myeloma among men stratified by race/ethnicity; and in situ melanoma by anatomic site among white women.

CONCLUSIONS

For studies of cancer rates with from two through to around 10 strata, which covers many outstanding questions in cancer surveillance research, our new method provides a comprehensive, coherent, and reproducible approach for joint analysis and synthesis of age-period-cohort estimable functions.

Trends and future projections of liver cancer incidence in Hong Kong: a population-based study

BACKGROUND

Liver cancer remains a significant burden in Hong Kong. We sought to examine trends in liver cancer incidence using 30 years of cancer registry data in Hong Kong. Additionally, we aimed to assess the impact of age, period and birth cohort on liver cancer incidence, provided projections for liver cancer incidence until 2030, and examined the drivers of liver cancer incidence.

METHODS

Data on liver cancer incidence were collected from the Hong Kong Cancer Registry (HKCaR). We assessed age, period, and birth cohort effects using age-period-cohort (APC) models. We employed Bayesian APC analysis with integrated nested Laplace approximations to project the future burden of liver cancer in Hong Kong. Furthermore, we attributed the changes in new liver cancer cases to population growth, population ageing, and epidemiological changes.

RESULTS

The study included a total of 51,333 individuals, of whom 39,287 (76.53%) were male. From 1991 to 2020, the age-standardized liver cancer incidence rate in Hong Kong continued declining, while the number of new cases increased significantly, especially among males. The net drift, representing the overall annual percentage change of the age-adjusted rate, was - 3.06% (95% confidence interval [CI]: -3.31% to -2.80%) for males and - 3.85% (95% CI: -4.61% to -3.09%) for females. Local drift, which estimates the annual percentage change over time specific to age group, decreased in all age groups for both sexes, with a more pronounced decrease in younger age groups. The period and cohort risk of developing liver cancer also showed decreasing trends for both sexes. The study projected a decline in liver cancer cases for males but an increase for females in Hong Kong, with an estimated 1,083 cases in males and 710 cases in females by 2030. Demographic decomposition analysis revealed that while population growth and ageing were the main drivers of increased liver cancer cases, epidemiologic shifts mostly offset these factors.

CONCLUSION

The period and cohort risk of developing liver cancer in Hong Kong declined due to epidemiological changes. Although the age-standardized incidence rates of liver cancer have also declined, demographic and epidemiological factors have led to lower case expectations in males but a likely increase in females. Further research and epidemiological assessment of the disease are needed.

Global trajectories of liver cancer burden from 1990 to 2019 and projection to 2035

BACKGROUND

Large disparities exist in liver cancer burden trends across countries but are poorly understood. We aimed to investigate the global trajectories of liver cancer burden, explore the driving forces, and predict future trends.

METHODS

Data on the liver cancer burden in 204 countries and territories from 1990 to 2019 were extracted from the Global Burden of Disease Study. The age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR) trajectories were defined using growth mixture models. Five major risk factors contributing to changes in the ASIR or ASMR and socioeconomic determinants were explored using the identified trajectories. A Bayesian age-period-cohort model was used to predict future trends through 2035.

RESULTS

Three trajectories of liver cancer burden were identified: increasing, stable, and decreasing groups. Almost half of the American countries were classified in the decreasing group (48.6% for ASIR and ASMR), and the increasing group was the most common in the European region (ASIR, 49.1%; ASMR, 37.7%). In the decreasing group, the decrease of liver cancer due to hepatitis B contributed 63.4% and 60.4% of the total decreases in ASIR and ASMR, respectively. The increase of liver cancer due to alcohol use, hepatitis C, and hepatitis B contributed the most to the increase in the increasing group (30.8%, 31.1%, and 24.2% for ASIR; 33.7%, 30.2%, and 22.2% for ASMR, respectively). The increasing group was associated with a higher sociodemographic index, gross domestic product per capita, health expenditure per capita, and universal health coverage (all P <0.05). Significant variations in disease burden are predicted to continue through 2035, with a disproportionate burden in the decreasing group.

CONCLUSION

Global disparities were observed in liver cancer burden trajectories. Hepatitis B, alcohol use, and hepatitis C were identified as driving forces in different regions.

Penalized smoothing splines resolve the curvature identifiability problem in age-period-cohort models with unequal intervals

Age-period-cohort (APC) models are frequently used in a variety of health and demographic-related outcomes. Fitting and interpreting APC models to data in equal intervals (equal age and period widths) is nontrivial due to the structural link between the three temporal effects (given two, the third can always be found) causing the well-known identification problem. The usual method for resolving the structural link identification problem is to base a model on identifiable quantities. It is common to find health and demographic data in unequal intervals, this creates further identification problems on top of the structural link. We highlight the new issues by showing that curvatures which were identifiable for equal intervals are no longer identifiable for unequal data. Furthermore, through extensive simulation studies, we show how previous methods for unequal APC models are not always appropriate due to their sensitivity to the choice of functions used to approximate the true temporal functions. We propose a new method for modeling unequal APC data using penalized smoothing splines. Our proposal effectively resolves the curvature identification issue that arises and is robust to the choice of the approximating function. To demonstrate the effectiveness of our proposal, we conclude with an application to UK all-cause mortality data from the Human mortality database.

Use of 2-Stage Modeling to Identify How Colorectal Cancer Risk Changes With Period and Cohort

Colorectal cancer (CRC) incidence rates have decreased among adults aged 50 years or older while increasing in adults under age 50 years. Understanding these trends is challenging because of the multiple related time scales of age, diagnosis period, and birth cohort. We analyzed incidence rates of rectal, distal colon, and proximal colon cancer for individuals aged 20 years or more from the Surveillance, Epidemiology, and End Results Program for diagnosis years 1978-2017. We used a 2-stage generalized linear model to determine age, period, and cohort effects for CRC incidence. We first estimated birth cohort effects among people under age 45 years. We used these results to specify prior distributions for cohort effects in a Bayesian model to estimate period effects among people aged 45 years or more. There was no evidence of period effects for people under age 45 years. Risks of rectal and distal colon cancer increased for later birth cohorts. Compared with the 1943-1952 birth cohort, the 1983-1992 birth cohort had 2.2 times the risk of rectal cancer, 1.9 times the risk of distal colon cancer, and 1.3 times the risk of proximal colon cancer. For people aged ≥45 years, period effects showed declines in CRC risk that were attributable to screening.

Secular trends of morbidity and mortality of prostate, bladder, and kidney cancers in China, 1990 to 2019 and their predictions to 2030

Background

Prostate, bladder and kidney cancers are common age-related genitourinary cancers. China's population is aging at an increasing rate, so predicting the morbidity and mortality of prostate, bladder, and kidney cancer in China is of great significance to provide epidemiological evidence for forward planning and implementation of national health policies.

Methods

Numbers of incidences and deaths by cancer (prostate, bladder and kidney), sex (male and female) and age groups from 1990 to 2019 were extracted from the Global Burden of Disease (GBD) Study. We applied Bayesian age-period-cohort models to predict incidences and deaths to 2030. We also calculated Age-standardized incidence rate (ASIR) and mortality rate (ASMR), their trends were quantified by estimated average percentage change (EAPC) and 95% confidence interval (CI).

Results

Predictions suggest that by 2030, there will be 315,310 prostate cancer cases, 192,390 bladder cancer cases and 126,980 kidney cancer cases. The ASIRs will increase to 25.54/100,000 for prostate cancer (EAPC: 2.88, 95% CI, 2.84, 2.93), 7.54/100,000 for bladder cancer (EAPC: 2.58, 95% CI, 2.54, 2.61) and 5.63/100,000 for kidney cancer (EAPC: 4.78, 95% CI, 4.54, 5.02). Number of deaths in 2030 will be 81,540, 61,220, and 41,940, respectively. Different ASMR changes are observed, the ASMR for prostate cancer will drop to 7.69/100,000 (EAPC: -0.29, 95% CI, -0.31, -0.27), the ASMR for bladder cancer will stabilize at 2.49/100,000 (EAPC: 0.00, 95% CI, -0.02, 0.03), the ASMR of kidney cancer will increase to 1.84/100,000 (EAPC: 3.45, 95% CI, 3.22, 3.67). From 1990 to 2030, higher numbers of cases and rates are reported among males and in the 60 plus age group, both ASIR and ASMR of bladder and kidney cancers presents progressively widening differences between both males and females and between the < 60 and the ≥ 60 age groups.

Conclusion

Morbidity and mortality of the three genitourinary cancers are predicted to increase further over the next decade. It highlights the need for timely development and implementation of optimal health policies to curb the epidemic trends.

Using arterial spin labeling to measure cerebrovascular reactivity in Moyamoya disease: Insights from simultaneous PET/MRI

Cerebrovascular reactivity (CVR) reflects the CBF change to meet different physiological demands. The reference CVR technique is PET imaging with vasodilators but is inaccessible to most patients. DSC can measure transit time to evaluate patients suspected of stroke, but the use of gadolinium may cause side-effects. Arterial spin labeling (ASL) is a non-invasive MRI technique for CBF measurements. Here, we investigate the effectiveness of ASL with single and multiple post labeling delays (PLD) to replace PET and DSC for CVR and transit time mapping in 26 Moyamoya patients. Images were collected using simultaneous PET/MRI with acetazolamide. CVR, CBF, arterial transit time (ATT), and time-to-maximum (Tmax) were measured in different flow territories. Results showed that CVR was lower in occluded regions than normal regions (by 68 ± 12%, 52 ± 5%, and 56 ± 9%, for PET, single- and multi-PLD PCASL, respectively, all p < 0.05). Multi-PLD PCASL correlated slightly higher with PET (CCC = 0.36 and 0.32 in affected and unaffected territories respectively). Vasodilation caused ATT to reduce by 4.5 ± 3.1% (p < 0.01) in occluded regions. ATT correlated significantly with Tmax (R² > 0.35, p < 0.01). Therefore, multi-PLD ASL is recommended for CVR studies due to its high agreement with the reference PET technique and the capability of measuring transit time.

Past, present, and future trends of overweight and obesity in Belgium using Bayesian age-period-cohort models

Background

Overweight and obesity are one of the most significant risk factors of the twenty-first century related to an increased risk in the occurrence of non-communicable diseases and associated increased healthcare costs. To estimate the future impact of overweight, the current study aimed to project the prevalence of overweight and obesity to the year 2030 in Belgium using a Bayesian age-period-cohort (APC) model, supporting policy planning.

Methods

Height and weight of 58,369 adults aged 18+ years, collected in six consecutive cross-sectional health interview surveys between 1997 and 2018, were evaluated. Criteria used for overweight and obesity were defined as body mass index (BMI) ≥ 25, and BMI ≥ 30. Past trends and projections were estimated with a Bayesian hierarchical APC model.

Results

The prevalence of overweight and obesity has increased between 1997 and 2018 in both men and women, whereby the highest prevalence was observed in the middle-aged group. It is likely that a further increase in the prevalence of obesity will be seen by 2030 with a probability of 84.1% for an increase in cases among men and 56.0% for an increase in cases among women. For overweight, it is likely to see an increase in cases in women (57.4%), while a steady state in cases among men is likely. A prevalence of 52.3% [21.2%; 83.2%] for overweight, and 27.6% [9.9%; 57.4%] for obesity will likely be achieved in 2030 among men. Among women, a prevalence of 49,1% [7,3%; 90,9%] for overweight, and 17,2% [2,5%; 61,8%] for obesity is most likely.

Conclusions

Our projections show that the WHO target to halt obesity by 2025 will most likely not be achieved. There is an urgent necessity for policy makers to implement effective prevent policies and other strategies in people who are at risk for developing overweight and/or obesity.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-022-13685-w.

Cancer incidence and mortality in Australia from 2020 to 2044 and an exploratory analysis of the potential effect of treatment delays during the COVID-19 pandemic: a statistical modelling study

BACKGROUND

Long-term projections of cancer incidence and mortality estimate the future burden of cancer in a population, and can be of great use in informing the planning of health services and the management of resources. We aimed to estimate incidence and mortality rates and numbers of new cases and deaths up until 2044 for all cancers combined and for 21 individual cancer types in Australia. We also illustrate the potential effect of treatment delays due to the COVID-19 pandemic on future colorectal cancer mortality rates.

METHODS

In this statistical modelling study, cancer incidence and mortality rates in Australia from 2020 to 2044 were projected based on data up to 2017 and 2019, respectively. Cigarette smoking exposure (1945-2019), participation rates in the breast cancer screening programme (1996-2019), and prostate-specific antigen testing rates (1994-2020) were included where relevant. The baseline projection model using an age-period-cohort model or generalised linear model for each cancer type was selected based on model fit statistics and validation with pre-COVID-19 observed data. To assess the impact of treatment delays during the COVID-19 pandemic on colorectal cancer mortality, we obtained data on incidence, survival, prevalence, and cancer treatment for colorectal cancer from different authorities. The relative risks of death due to system-caused treatment delays were derived from a published systematic review. Numbers of excess colorectal cancer deaths were estimated using the relative risk of death per week of treatment delay and different durations of delay under a number of hypothetical scenarios.

FINDINGS

Projections indicate that in the absence of the COVID-19 pandemic effects, the age-standardised incidence rate for all cancers combined for males would decline over 2020-44, and for females the incidence rate would be relatively stable in Australia. The mortality rates for all cancers combined for both males and females are expected to continuously decline during 2020-44. The total number of new cases are projected to increase by 47·4% (95% uncertainty interval [UI] 35·2-61·3) for males, from 380 306 in 2015-19 to 560 744 (95% UI 514 244-613 356) in 2040-44, and by 54·4% (95% UI 40·2-70·5) for females, from 313 263 in 2015-19 to 483 527 (95% UI 439 069-534 090) in 2040-44. The number of cancer deaths are projected to increase by 36·4% (95% UI 15·3-63·9) for males, from 132 440 in 2015-19 to 180 663 (95% UI 152 719-217 126) in 2040-44, and by 36·6% (95% UI 15·8-64·1) for females, from 102 103 in 2015-19 to 139 482 (95% UI 118 186-167 527) in 2040-44, due to population ageing and growth. The example COVID-19 pandemic scenario of a 6-month health-care system disruption with 16-week treatment delays for colorectal cancer patients could result in 460 (95% UI 338-595) additional deaths and 437 (95% UI 314-570) deaths occurring earlier than expected in 2020-44.

INTERPRETATION

These projections can inform health service planning for cancer care and treatment in Australia. Despite the continuous decline in cancer mortality rates, and the decline or plateau in incidence rates, our projections suggest an overall 51% increase in the number of new cancer cases and a 36% increase in the number of cancer deaths over the 25-year projection period. This means that continued efforts to increase screening uptake and to control risk factors, including smoking exposure, obesity, physical inactivity, alcohol use, and infections, must remain public health priorities.

FUNDING

Partly funded by Cancer Council Australia.

Trends in colon and rectal cancer mortality in Australia from 1972 to 2015 and associated projections to 2040

Previously published sub-site Australian projections for colon and rectal cancers to 2035 using the World Health Organization's mortality database sourced from the Australian Bureau of Statistics (ABS) predicted mortality rate decreases for colon cancer and increases for rectal cancer. There are complexities related to the interpretation of ABS's Australian colon and rectal cancer mortality rates, which could lead to possible inaccuracies in mortality rates for these sub-sites. The largest Australian population-wide registry, New South Wales Cancer Registry (NSWCR), compares routinely-reported causes of death with the recorded medical history from multiple data sources. Therefore, this study used the NSWCR data to project mortality rates for colon and rectal cancers separately to 2040 in Australia. The mortality rates for colon cancer are projected to continuously decline over the period 2015-2040, from 7.0 to 4.7 per 100,000 males, and from 5.3 to 3.2 per 100,000 females. Similar decreasing trends in mortality rates for rectal cancer were projected over the period 2015-2040, from 4.9 to 3.7 per 100,000 males, and from 2.6 to 2.3 per 100,000 females. These projections provide benchmark estimates for the colorectal cancer burden in Australia against which the effectiveness of cancer control interventions can be measured.

Secular Trends in the Incidence of Migraine in China from 1990 to 2019: A Joinpoint and Age–Period–Cohort Analysis

Background

Migraine is a common disorder of the nervous system in China, imposing heavy burdens on individual and societies. Optimal healthcare planning requires understanding the magnitude and changing the trend of migraine incidence in China. However, the secular trend of migraine incidence in China remains unclear.

Methods

Data were collected from the Global Burden of Disease Study 2019 in China from 1990 to 2019 to investigate changes in the incidence rate of migraine. The average annual percent change and the relative risk were calculated using the joinpoint regression model and the age–period–cohort model, respectively.

Results

From 1990 to 2019, the age-standardized incidence rates of migraine in China increased by 0.26% (95% CI: 0.22 to 0.31) and 0.23% (95% CI: 0.19 to 0.28) per year in males and females, respectively. Age effects exerted the most significant impact on migraine incidence. Period effects showed a slightly decreasing trend in the incidence of migraine. In terms of cohort effects, people born after the 1960s presented a higher risk of migraine as compared with the total cohort, with the incidence risk of migraine increasing with birth cohorts.

Conclusion

Migraine incidence shows an overall increasing trend in China, with a significant gender difference. A comprehensive understanding of the risk characteristics and disease pattern of migraine could allow the early detection of persons with a high risk of developing migraine and promote the development of timely intervention measures to relieve this burden effectively.

Trends and age-period-cohort effects on hypertension mortality rates from 1998 to 2018 in Mexico

Arterial hypertension is a major global health problem. It is the main risk factor for preventable death and the leading cause of premature death in the world. This study aims to describe the changes in hypertension-related mortality in Mexico between 1998 and 2018. Using death certificates and national population public data sets, a total of 335,863 deaths due to hypertension were found in Mexico, disaggregated by sex and age, during the time period covered in this study. An age-period-cohort analysis was conducted to show trends in hypertension mortality rates. Mortality due to hypertension in Mexico affects more women than men. In the most recent cohorts, the risk of dying from hypertension is two times higher in men compared to women. Hypertensive kidney disease is found to be the main underlying cause, with an average increase throughout the period studied. Our results indicate that mortality rates due to hypertension continue to grow and point to an alarming trend of mortality shifting towards younger ages, with sex-based disparities in absolute numbers and in changing trends.

Cigarette smoking trajectories in adolescent smokers: Does the time axis metric matter?

INTRODUCTION

Most studies modeling adolescent cigarette smoking trajectories use age as the time axis, possibly obscuring depiction of the natural course of cigarette smoking. We used a simulated example and real data to contrast smoking trajectories obtained from models that used time since smoking onset or calendar time (age) as the time axis.

METHODS

Data were drawn from a longitudinal investigation of 1293 grade 7 students (mean age 12.8 years) recruited from 10 high schools in Montreal, Canada in 1999-2000, who were followed into young adulthood. Cigarette consumption was measured every 3 months during high school, and again at mean age 20.4 and 24.0. Analyses using time since onset of smoking as the time metric was restricted to 307 incident smokers; analysis using calendar time included 645 prevalent and incident smokers. Smoking status and nicotine dependence (ND) were assessed at mean age 20.4 and 24.0. Simulated data mimicked the real study during high school.

RESULTS

Use of different time metrics resulted in different numbers and shapes of trajectories in the simulated and real datasets. Participants in the calendar time analyses reported more ND in young adulthood, reflecting inclusion of 388 prevalent smokers who had smoked for longer durations.

CONCLUSIONS

Choosing the right time metric for trajectory analysis should be balanced against research intent. Trajectory analyses using the time since onset metric depict the natural course of smoking in incident smokers. Those using calendar time offer a snapshot of smoking across age during a given time period.

IMPLICATIONS

This study uses simulated and real data to show that trajectory analyses of cigarette smoking that use calendar time (e.g., age) versus time since onset as the time axis metric tell a different story. Trajectory analyses using the time since onset metric depict the natural course of smoking in incident smokers. Those using calendar time offer a snapshot of smoking across age during a given time period. Choosing the right time metric should be balanced against research intent.

Global Patterns and Trends in Gastric Cancer Incidence Rates (1988-2012) and Predictions to 2030

BACKGROUND & AIMS

The long-term trend in gastric cancer rates has rarely been reported from a global perspective. We aimed to explore the past temporal trends (1988-2012) in gastric cancer incidence rates in 43 countries and to predict future trends (2012-2030).

METHODS

Data on yearly gastric cancer incidence by age group and sex were drawn from 108 cancer registries in 43 countries in the Cancer Incidence in Five Continents Time Trends (CI5plus) database. Age-standardized incidence rates per 100,000 persons were computed from 1988-2012. The number of new cases and incidence rates were predicted to 2030 using the Bayesian age-period-cohort model.

RESULTS

Persistent decreasing trends in gastric cancer incidence rates were observed from 1988-2012 worldwide, with an overall average annual percentage change of -2.1% (95% confidence interval, -2.5--1.7). The trends will continue or remain stable until 2030 in most of the selected countries except for Ecuador and Lithuania, whose gastric cancer incidence rates will experience substantially increasing trends in the next several decades.

CONCLUSION

The incidence rates of gastric cancer are expected to decrease through 2030 in most countries except Ecuador and Lithuania. Reductions in smoking and Helicobacter pylori prevalence and improvement in diet probably contributed to the decrease. Gastric cancer still represents a major cancer burden worldwide, and the large number of gastric cancer cases worldwide may still call for lifestyle interventions in terms of smoking and diet and massive efforts for H pylori screening and treatment, especially in countries with predicted increasing incidence rates of gastric cancer.

An Analysis of the Trend of Fetal Mortality Rates among Working and Jobless Households in Japan, 1995-2019

This study aimed to identify differences in the trends of artificial and spontaneous fetal mortality rates between working and jobless households depending on ages, periods, and birth cohorts in Japan. Vital Statistics data from 1995 to 2019 and age groups in 5–year increments from 15 to 19 years through 45 to 49 years were used. Bayesian age–period–cohort analysis was used to evaluate changes in each of the outcomes. As a result, the difference in maternal age–standardized rate of both the artificial and spontaneous fetal mortality rates between the two types of households decreased in the periods analyzed. However, there was a statistically significant difference in the mortality rate between jobless and working households, regardless of maternal ages, periods, and cohorts for the artificial fetal mortality rate. A statistically significant difference was also observed for the spontaneous fetal mortality rates in some maternal ages, periods, and cohorts. In addition, the trend of birth cohort effects was particularly different between the two types of households for both the artificial and spontaneous fetal mortality rates.

Multidimensional penalized splines for incidence and mortality-trend analyses and validation of national cancer-incidence estimates

BACKGROUND

Cancer-incidence and mortality-trend analyses require appropriate statistical modelling. In countries without a nationwide cancer registry, an additional issue is estimating national incidence from local-registry data. The objectives of this study were to (i) promote the use of multidimensional penalized splines (MPS) for trend analyses; (ii) estimate the national cancer-incidence trends, using MPS, from only local-registry data; and (iii) propose a validation process of these estimates.

METHODS

We used an MPS model of age and year for trend analyses in France over 1990-2015 with a projection up to 2018. Validation was performed for 22 cancer sites and relied essentially on comparison with reference estimates that used the incidence/health-care ratio over the period 2011-2015. Alternative estimates that used the incidence/mortality ratio were also used to validate the trends.

RESULTS

In the validation assessment, the relative differences of the incidence estimates (2011-2015) with the reference estimates were <5% except for testis cancer in men and < 7% except for larynx cancer in women. Trends could be correctly derived since 1990 despite incomplete histories in some registries. The proposed method was applied to estimate the incidence and mortality trends of female lung cancer and prostate cancer in France.

CONCLUSIONS

The validation process confirmed the validity of the national French estimates; it may be applied in other countries to help in choosing the most appropriate national estimation method according to country-specific contexts. MPS form a powerful statistical tool for trend analyses; they allow trends to vary smoothly with age and are suitable for modelling simple as well as complex trends thanks to penalization. Detailed trend analyses of lung and prostate cancers illustrated the suitability of MPS and the epidemiological interest of such analyses.

Incidence of Benign Meningiomas in the United States: Current and Future Trends

Background

Benign meningiomas are the most frequently reported central nervous system tumors in the United States, with increasing incidence in past decades. However, the future trajectory of this neoplasm remains unclear.

Methods

We analyzed benign meningioma incidence of cases identified by any means (eg, radiographically with or without microscopic confirmation) in US Surveillance, Epidemiology, and End Results cancer registries among groups aged 35 to 84 years during 2004-2017 by sex and race and ethnicity using age-period-cohort models. We employed age-period-cohort forecasting models to glean insights regarding the etiology, distribution, and anticipated future (2018-2027) public health impact of this neoplasm.

Results

In all groups, meningioma incidence overall increased through 2010, then stabilized. Temporal declines were statistically significant overall and in most groups. JoinPoint analysis of cohort rate-ratios identified substantial acceleration in White men born after 1963 (from 1.1% to 3.2% per birth year); cohort rate-ratios were stable or increasing in all groups and all birth cohorts. We forecast that meningioma incidence through 2027 will remain stable or decrease among groups aged 55-84 years but remain similar to current levels among groups aged 35-54 years. The case count of total meningioma burden in 2027 is expected to be approximately 30 470, similar to the expected case count of 27 830 in 2018.

Conclusions

Between 2004 and 2017, overall incidence of benign meningioma increased and then stabilized or declined. For 2018-2027, our forecast is incidence will remain generally stable in younger age groups but decrease in older age groups. Nonetheless, the total future burden will remain similar to current levels because the population is aging.

Projections of the future burden of cancer in Australia using Bayesian age-period-cohort models

BACKGROUND

Accurate forecasts of cancer incidence, with appropriate estimates of uncertainty, are crucial for planners and policy makers to ensure resource availability and prioritize interventions. We used Bayesian age-period-cohort (APC) models to project the future incidence of cancer in Australia.

METHODS

Bayesian APC models were fitted to counts of cancer diagnoses in Australia from 1982 to 2016 and projected to 2031 for seven key cancer types: breast, colorectal, liver, lung, non-Hodgkin lymphoma, melanoma and stomach. Aggregate cancer data from population-based cancer registries were sourced from the Australian Institute of Health and Welfare.

RESULTS

Over the projection period, total counts for these cancer types increased on average by 3 % annually to 100 385 diagnoses in 2031, which is a 50 % increase over 2016 numbers, although there is considerable uncertainty in this estimate. Counts for each cancer type and sex increased over the projection period, whereas decreases in the age-standardized incidence rates (ASRs) were projected for stomach, colorectal and male lung cancers. Large increases in ASRs were projected for liver and female lung cancer. Increases in the percentage of colorectal cancer diagnoses among younger age groups were projected. Retrospective one-step-ahead projections indicated both the incidence and its uncertainty were successfully forecast.

CONCLUSIONS

Increases in the projected incidence counts of key cancer types are in part attributable to the increasing and ageing population. The projected increases in ASRs for some cancer types should increase motivation to reduce sedentary behaviour, poor diet, overweight and undermanagement of infections. The Bayesian paradigm provides useful measures of the uncertainty associated with these projections.

An age-period-cohort analysis for prevalence of common psychiatric disorders in Japan, 1999-2017

PURPOSE

This study aimed to analyze the prevalence of common psychiatric disorders in Japan via an age-period-cohort (APC) analysis.

METHODS

Data were used on the prevalence of schizophrenia, schizotypal, and delusional disorders; mood [affective] disorders (mood disorders, hereafter); and neurotic, stress-related, and somatoform disorders from the 1999-2017 Patient Survey in Japan. The age group was defined as 20-89 years with 5-year increments; the cohort was assigned for each age group of each year with a 1-year shift. A Bayesian APC analysis was used to decompose changes in prevalence into three effects: age, period, and cohort.

RESULTS

The APC analysis revealed that the peaks of age effect varied among sexes and diseases. The period effects for all the diseases showed increasing trends over the analyzed years. Although the cohort effect for the prevalence of schizophrenia, schizotypal, and delusional disorders increased until the 1960s, it decreased afterwards. On the other hand, the cohort effects for the prevalence of mood disorders and neurotic, stress-related, and somatoform disorders increased from around the 1950s, and it indicates the prevalence increased particularly in young ages over the years. Also, the trends of each effect were relatively similar between mood disorders and neurotic, stress-related, and somatoform disorders.

CONCLUSION

Increase in public awareness and psychological stress associated with a change in a social environment is thought to affect the period and cohort effects on the prevalence of mood disorders and neurotic, stress-related, and somatoform disorders.

Burden of Thyroid Cancer From 1990 to 2019 and Projections of Incidence and Mortality Until 2039 in China: Findings From Global Burden of Disease Study

Projecting the burden of thyroid cancer (TC) over time provides essential information to effectively plan measures for its management and prevention. This research obtained data from the Global Burden of Disease (GBD) Study from between 1990 and 2019 to model how TC will affect China until 2039 by conducting the Bayesian age-period-cohort analysis. The number of new TC cases in China was 10,030 in 1990, 39,080 in 2019, and is projected to be 47,820 in 2039. This corresponds to 3,320, 7,240, and 4,160 deaths, respectively. Disability-adjusted life years (DALYs) cases increased from 103,490 in 1990 to 187,320 in 2019. The age-standardized rate (ASR) of incidence increased from 1.01 to 2.05 during 1990-2019 and was projected to increase to 3.37 per 100,000 person-years until 2039. The ASR of mortality (ASMR) remained stable during the study period and was projected to have a mild decline from 0.39 to 0.29/100,000 during 2020-2039. Although the ASMR in male patients has maintained increasing at a rate of 2.2% per year over the past 30 years, it is expected to decline at a rate of 1.07% per year in 2019-2039. The most significant increase in crude incidence occurred in people aged 45-65 from 1990 to 2019, however, this will shift into young people aged 10-24 from 2020 to 2039. In addition, the proportion of deaths and DALYs caused by obesity increased from 1990 to 2019 and affected men more than women. In conclusion, a substantial increase in counts of incidence of TC in China is projected over the next two decades, combined with the slightly declining mortality, indicating that rational health policies are needed in the future to cope with the increasing number of TC patients, especially among males and adolescents.

An age-period-cohort analysis of mortality rates for stomach, colorectal, liver, and lung cancer among prefectures in Japan, 1999-2018

BACKGROUND

Although change in the birth cohort effect on cancer mortality rates is known to be highly associated with the decreasing rates of age-standardized cancer mortality rates in Japan, the differences in the trends of cohort effect for representative cancer types among the prefectures remain unknown. This study aimed to investigate the differences in the decreasing rate of cohort effects among the prefectures for representative cancer types using age-period-cohort (APC) analysis.

METHODS

Data on stomach, colorectal, liver, and lung cancer mortality for each prefecture and the population data from 1999 to 2018 were obtained from the Vital Statistics in Japan. Mortality data for individuals aged 50 to 79 years grouped in 5-year increments were used, and corresponding birth cohorts born 1920-1924 through 1964-1978 were used for analysis. We estimated the effects of age, period, and cohort on each type of mortality rate for each prefecture by sex. Then, we calculated the decreasing rates of cohort effects for each prefecture. We also calculated the mortality rate ratio of each prefecture compared with all of Japan for cohorts using the estimates.

RESULTS

As a result of APC analysis, we found that the decreasing rates of period effects were small and that there was a little difference in the decreasing rates among prefectures for all types of cancer among both sexes. On the other hand, there was a large difference in the decreasing rates of cohort effects for stomach and liver cancer mortality rates among prefectures, particularly for men. For men, the decreasing rates of cohort effects in cohorts born between 1920-1924 and 1964-1978 varied among prefectures, ranging from 4.1 to 84.0% for stomach cancer and from 20.2 to 92.4% for liver cancers, respectively. On the other hand, the differences in the decreasing rates of cohort effects among prefectures for colorectal and lung cancer were relatively smaller.

CONCLUSIONS

The decreasing rates of cohort effects for stomach and liver cancer varied widely among prefectures. It is possible that this will influence cancer mortality rates in each prefecture in the future.

An Age-Period-Cohort Analysis of Biomarkers of Lifestyle-Related Diseases Using the National Health and Nutrition Survey in Japan, 1973-2018

Studies of biomarkers of lifestyle-related diseases in Japanese cohorts are scarce. This study aimed to analyze trends in risk markers of lifestyle-related diseases using age-period-cohort (APC) analysis. Data on systolic blood pressure and BMI from 1973 to 2018 and serum glucose, triglyceride, and high-density lipoprotein cholesterol levels from 1989 to 2018 available from the National Health and Nutrition Survey were used. Values for each of the risk markers for each age, period, and cohort were estimated using APC analysis. For women, a decrease in all the risk markers of lifestyle-related diseases was observed in individuals born between the 1930s and approximately 1970. Therefore, female individuals born in approximately 1970 were considered to have the lowest risk of developing lifestyle-related and cardiovascular diseases. Meanwhile, the cohort effect on all the risk markers deteriorated for the younger cohorts, and changes in lifestyle behavior are needed for cohorts born more recently. For men, the trends in risk markers across the cohorts differed, and the relative risk of lifestyle-related diseases for each cohort differed according to disease. These results could help understand cohort-specific risks for lifestyle-related disease and enable identification of high-risk populations who could benefit from preventive measures.

Prenatal and early-life exposure to the Great Chinese Famine increased the risk of tuberculosis in adulthood across two generations

Global food security is a major driver of population health, and food system collapse may have complex and long-lasting effects on health outcomes. We examined the effect of prenatal exposure to the Great Chinese Famine (1958-1962)-the largest famine in human history-on pulmonary tuberculosis (PTB) across consecutive generations in a major center of ongoing transmission in China. We analyzed >1 million PTB cases diagnosed between 2005 and 2018 in Sichuan Province using age-period-cohort analysis and mixed-effects metaregression to estimate the effect of the famine on PTB risk in the directly affected birth cohort (F1) and their likely offspring (F2). The analysis was repeated on certain sexually transmitted and blood-borne infections (STBBI) to explore potential mechanisms of the intergenerational effects. A substantial burden of active PTB in the exposed F1 cohort and their offspring was attributable to the Great Chinese Famine, with more than 12,000 famine-attributable active PTB cases (>1.23% of all cases reported between 2005 and 2018). An interquartile range increase in famine intensity resulted in a 6.53% (95% confidence interval [CI]: 1.19-12.14%) increase in the ratio of observed to expected incidence rate (incidence rate ratio, IRR) in the absence of famine in F1, and an 8.32% (95% CI: 0.59-16.6%) increase in F2 IRR. Increased risk of STBBI was also observed in F2. Prenatal and early-life exposure to malnutrition may increase the risk of active PTB in the exposed generation and their offspring, with the intergenerational effect potentially due to both within-household transmission and increases in host susceptibility.

An R package for an integrated evaluation of statistical approaches to cancer incidence projection

Background

Projection of future cancer incidence is an important task in cancer epidemiology. The results are of interest also for biomedical research and public health policy. Age-Period-Cohort (APC) models, usually based on long-term cancer registry data (> 20 yrs), are established for such projections. In many countries (including Germany), however, nationwide long-term data are not yet available. General guidance on statistical approaches for projections using rather short-term data is challenging and software to enable researchers to easily compare approaches is lacking.

Methods

To enable a comparative analysis of the performance of statistical approaches to cancer incidence projection, we developed an R package (incAnalysis), supporting in particular Bayesian models fitted by Integrated Nested Laplace Approximations (INLA). Its use is demonstrated by an extensive empirical evaluation of operating characteristics (bias, coverage and precision) of potentially applicable models differing by complexity. Observed long-term data from three cancer registries (SEER-9, NORDCAN, Saarland) was used for benchmarking.

Results

Overall, coverage was high (mostly > 90%) for Bayesian APC models (BAPC), whereas less complex models showed differences in coverage dependent on projection-period. Intercept-only models yielded values below 20% for coverage. Bias increased and precision decreased for longer projection periods (> 15 years) for all except intercept-only models. Precision was lowest for complex models such as BAPC models, generalized additive models with multivariate smoothers and generalized linear models with age x period interaction effects.

Conclusion

The incAnalysis R package allows a straightforward comparison of cancer incidence rate projection approaches. Further detailed and targeted investigations into model performance in addition to the presented empirical results are recommended to derive guidance on appropriate statistical projection methods in a given setting.

An Age-Period-Cohort Analysis of Prevalence and Consultation Rate for Dyslipidemia in Japan

This study involved an age-period-cohort analysis of the consultation rate and prevalence of dyslipidemia in Japan, based on Patient Survey data from 1999 to 2017 and open data of national database of health insurance claims and specific health checkups in Japan from 2013 to 2016. Our results showed that the consultation rates were lower than the prevalence, regardless of age, year, and sex, and particularly among middle-aged and male respondents. Additionally, both the consultation rate and prevalence increased with increasing age to a greater extent among women than men, and the degree of increase in the consultation rate was larger than that in prevalence among women. Furthermore, although the cohort effect on prevalence began to decrease among men in cohorts born in approximately 1960, the effect decreased among women in cohorts born between the 1930s and 1960s and exhibited an increasing trend thereafter.

Spatially varying age-period-cohort analysis with application to US mortality, 2002-2016

Many public health databases index disease counts by age groups and calendar periods within geographic regions (e.g., states, districts, or counties). Issues around relative risk estimation in small areas are well-studied; however, estimating trend parameters that vary across geographic regions has received less attention. Additionally, small counts (e.g., $<10$) in most publicly accessible databases are censored, further complicating age-period-cohort (APC) analysis in small areas. Here, we present a novel APC model with left-censoring and spatially varying intercept and trends, estimated with correlations among contiguous geographic regions. Like traditional models, our model captures population-scale trends, but it can also be used to characterize geographic disparities in relative risk and age-adjusted trends over time. To specify the joint distribution of our three spatially varying parameters, we adapt the generalized multivariate conditional autoregressive prior, previously used for multivariate disease mapping. Specified in this manner, region-specific parameters are correlated spatially, and also to one another. Estimation is performed using the No-U-Turn Hamiltonian Monte Carlo sampler in Stan. We conduct a simulation study to assess the performance of the proposed model relative to the standard model, and conclude with an application to US state-level opioid overdose mortality in men and women aged 15-64 years.

Age-period-cohort Analysis of Healthy Lifestyle Behaviors Using the National Health and Nutrition Survey in Japan

OBJECTIVES

This study conducted an age-period-cohort (APC) analysis of trends in healthy lifestyle behaviors in Japan.

METHODS

We used National Health and Nutrition Survey data on salt intake and prevalence of smoking, drinking, and physical activity between 1995 and 2018 in Japan. Age groups were defined from 20 years to 69 years old in 10-year increments. Cohorts were defined for each age group of each year with a 1-year shift, and cohorts born in 1926-1935 (first cohort) until 1989-1998 (last cohort) were examined. We conducted a Bayesian APC analysis, calculating estimated values for each behavior by age group, period, and cohort.

RESULTS

Estimated salt intake decreased from cohorts born in the 1930s to the 1960s, but increased thereafter in both genders, and the magnitude of increase was larger for men. Estimated smoking prevalence increased in the cohorts starting from the 1930s for men and the 1940s for women, and then decreased starting in the cohorts born in the 1970s for both genders. Although estimated drinking prevalence decreased starting in the cohorts born in approximately 1960 for men, for women it increased until the cohorts born in approximately 1970. Estimated physical activity prevalence decreased starting in the cohorts born in the 1940s in both genders, but the magnitude of decrease was larger for women.

CONCLUSIONS

Trends in cohort effects differed by gender, which might be related to changes in the social environment for women. Improvements in dietary and exercise habits are required in more recently born cohorts of both genders.

Age-period-cohort analysis of asthma, allergic rhinitis, and atopic dermatitis prevalence in Japan

This study aims to analyze the trends in the Japanese prevalence of asthma, allergic rhinitis, and atopic dermatitis by using age–period–cohort (APC) analysis. Data regarding the prevalence of diseases from 1999 to 2017 were collected from Patient Survey in Japan. The data were divided according to age groups ranging from 0–4 years old up to 65–69 years old in 5-year increments. A cohort was defined for each age group of each year with a one-year shift, and cohorts born from 1930–1934 up to 2013–2017 were examined. We used Bayesian APC analysis to decompose the changes in prevalence into age, period, and cohort effects. Results show that the period effect for asthma began to increase in 2008, and those of allergic rhinitis and atopic dermatitis began to increase in 1999. The cohort effects for asthma and atopic dermatitis increased rapidly in cohorts born from approximately 1950 to 1980 and then decreased thereafter. Furthermore, the cohort effect for allergic rhinitis increased from cohorts born in approximately the late 1970s for men and in 1990 for women. The time points with increasing cohort effects for asthma and atopic dermatitis are consistent with the history of air pollution accompanied by rapid economic growth in Japan. The onset of the increased cohort effect for allergic rhinitis was also relatively consistent with the time point at which the mass scattering of pollen began.

Age-period-cohort Analysis of Cardiovascular Disease Mortality in Japan, 1995-2018

OBJECTIVES

This study aimed to analyze the mortality of heart disease (HD), ischemic heart disease (IHD), and cerebrovascular disease (CeVD) through an age-period-cohort (APC) analysis.

METHODS

We used data on mortality due to cardiovascular disease from 1995 to 2018 in Japan, as determined by Vital Statistics. Age groups from 0 years to 99 years were defined by 5-year increments, and cohorts were defined for each age group of each year with a 1-year shift. We used Bayesian APC analysis to decompose the changes in the diseases' mortality rates into age, period, and cohort effects.

RESULTS

The period effects for all diseases decreased during the analyzed periods for both men and women. The cohort effects for men increased substantially in cohorts born from around 1940 to the 1970s for all types of cardiovascular diseases. The cohort effects of HD decreased in the cohorts born in the 1970s or later for both men and women. Regarding IHD and CeVD, either a non-increase or decrease of cohort effects was confirmed for cohorts born in the 1970s or later for men, but the effects for women showed a continuously increasing trend in the cohorts born in the 1960s or later.

CONCLUSIONS

The cohort effects for IHD and CeVD showed increasing trends in younger generations of women. This suggests that preventive approaches against cardiovascular diseases are needed, particularly for women.

Pain rates in general population for the period 1991-2015 and 10-years prediction: results from a multi-continent age-period-cohort analysis

Background

Pain is a common symptom, often associated with neurological and musculoskeletal conditions, and experienced especially by females and by older people. The aims of this study are to evaluate the temporal variations of pain rates among general populations for the period 1991–2015 and to project 10-year pain rates.

Methods

We used the harmonized dataset of ATHLOS project, which included 660,028 valid observations in the period 1990–2015 and we applied Bayesian age–period–cohort modeling to perform projections up to 2025. The harmonized Pain variable covers the content “self-reported pain experienced at the time of the interview”, with a dichotomous (yes or no) modality.

Results

Pain rates were higher among females, older subjects, in recent periods, and among observations referred to cohorts of subjects born between the 20s and the 60s. The 10-year projections indicate a noteworthy increase in pain rates in both genders and particularly among subjects aged 66 or over, for whom a 10–20% increase in pain rate is foreseen; among females only, a 10–15% increase in pain rates is foreseen for those aged 36–50.

Conclusions

Projected increase in pain rates will require specific interventions by health and welfare systems, as pain is responsible for limited quality of subjective well-being, reduced employment rates and hampered work performance. Worksite and lifestyle interventions will therefore be needed to limit the impact of projected higher pain rates.

Bounding Analyses of Age-Period-Cohort Effects

For more than a century, researchers from a wide range of disciplines have sought to estimate the unique contributions of age, period, and cohort (APC) effects on a variety of outcomes. A key obstacle to these efforts is the linear dependence among the three time scales. Various methods have been proposed to address this issue, but they have suffered from either ad hoc assumptions or extreme sensitivity to small differences in model specification. After briefly reviewing past work, we outline a new approach for identifying temporal effects in population-level data. Fundamental to our framework is the recognition that it is only the slopes of an APC model that are unidentified, not the nonlinearities or particular combinations of the linear effects. One can thus use constraints implied by the data along with explicit theoretical claims to bound one or more of the APC effects. Bounds on these parameters may be nearly as informative as point estimates, even with relatively weak assumptions. To demonstrate the usefulness of our approach, we examine temporal effects in prostate cancer incidence and homicide rates. We conclude with a discussion of guidelines for further research on APC effects.

ESTIMATING AND FORECASTING THE SMOKING-ATTRIBUTABLE MORTALITY FRACTION FOR BOTH GENDERS JOINTLY IN OVER 60 COUNTRIES

Smoking is one of the leading preventable threats to human health and a major risk factor for lung cancer, upper aero-digestive cancer, and chronic obstructive pulmonary disease. Estimating and forecasting the smoking attributable fraction (SAF) of mortality can yield insights into smoking epidemics and also provide a basis for more accurate mortality and life expectancy projection. Peto et al. (1992) proposed a method to estimate the SAF using the lung cancer mortality rate as an indicator of exposure to smoking in the population of interest. Here we use the same method to estimate the all-age SAF (ASAF) for both genders for over 60 countries. We document a strong and cross-nationally consistent pattern of the evolution of the SAF over time. We use this as the basis for a new Bayesian hierarchical model to project future male and female ASAF from over 60 countries simultaneously. This gives forecasts as well as predictive distributions that can be used to find uncertainty intervals for any quantity of interest. We assess the model using out-of-sample predictive validation, and find that it provides good forecasts and well calibrated forecast intervals, comparing favorably with other methods.

Age-Period-Cohort Analysis on the Time Trend of Hepatitis B Incidence in Four Prefectures of Southern Xinjiang, China from 2005 to 2017

Objective: The influence of age, period, and cohort on Hepatitis B (HB) incidence in four prefectures of southern Xinjiang, China is still not clear. This paper aims to analyze the long-term trend of the HB incidence in four prefectures of southern Xinjiang, China and to estimate the independent impact of age, period and cohort, as well as to predict the development trend of HB incidence in male and female groups, then to identify the targeted population for HB screening by the model fitting and prediction. Method: The data were from the Case List of HB Cases Reported in the Infectious Disease Reporting Information Management System and the Xinjiang Statistical Yearbook of China. The age-period-cohort (APC) model was used to estimate the impacts of age, period and cohort on HB incidence, which could be used to predict the HB incidence in specific age groups of men and women. Results: Under the influence of age effect, the incidence of HB in males had two peaks (20–35 years old and 60–80 years old), the influence of age effect on the incidence of HB in females was lower than that of males and the obvious peak was between 20–30 years old; the period effect on the HB incidence in males and females fluctuated greatly and the fluctuation degree of influence on males was bigger than that of women. The HB incidence among males and females in the four regions tended to be affected by cohort effect, which reached a peak after 1990 and then declined sharply and gradually became stabilized. By predicting the HB incidence from 2018 to 2022, we found that there were significant differences in HB incidence among people over 35 years old, under 35 years old and the whole population in four prefectures of southern Xinjiang, China. Conclusions: Although the incidence of HB in some regions shows a downward trend, there is still an obvious upward trend of incidences in other places. In our paper, results indicate that the burden of HB incidence may be extended in the future, so we hope this can draw the attention of relative departments. These results reveal the differences of incidence between males and females as well, so respective measures of the two groups’ functions are essential.

Statistical projection methods for lung cancer incidence and mortality: a systematic review

OBJECTIVES

To identify and summarise all studies using statistical methods to project lung cancer incidence or mortality rates more than 5 years into the future.

STUDY TYPE

Systematic review.

METHODS

We performed a systematic literature search in multiple electronic databases to identify studies published from 1 January 1988 to 14 August 2018, which used statistical methods to project lung cancer incidence and/or mortality rates. Reference lists of relevant articles were checked for additional potentially relevant articles. We developed an organisational framework to classify methods into groups according to the type of data and the statistical models used. Included studies were critically appraised using prespecified criteria.

RESULTS

One hundred and one studies met the inclusion criteria; six studies used more than one statistical method. The number of studies reporting statistical projections for lung cancer increased substantially over time. Eighty-eight studies used projection methods, which did not incorporate data on smoking in the population, and 16 studies used a method which did incorporate data on smoking. Age-period-cohort models (44 studies) were the most commonly used methods, followed by other generalised linear models (35 studies). The majority of models were developed using observed rates for more than 10 years and used data that were considered to be good quality. A quarter of studies provided comparisons of fitted and observed rates. While validation by withholding the most recent observed data from the model and then comparing the projected and observed rates for the most recent period provides important information on the model's performance, only 12 studies reported doing this.

CONCLUSION

This systematic review provides an up-to-date summary of the statistical methods used in published lung cancer incidence or mortality projections. The assessment of the strengths of existing methods will help researchers to better apply and develop statistical methods for projecting lung cancer rates. Some of the common methods described in this review can be applied to the projection of rates for other cancer types or other non-infectious diseases.

Lung Cancer Mortality Trends in a Brazilian City with a Long History of Asbestos Consumption

There are scarce epidemiological studies on lung cancer mortality in areas exposed to asbestos in developing countries. We compared the rates and trends in mortality from lung cancer between 1980 and 2016 in a municipality that made extensive use of asbestos, Osasco, with rates from a referent municipality with lower asbestos exposure and with the rates for the State of São Paulo. We retrieved death records for cases of lung cancer (ICD-9 C162) (ICD-10 C33 C34) from 1980 to 2016 in adults aged 60 years and older. The join point regression and age-period-cohort models were fitted to the data. Among men, there was an increasing trend in lung cancer mortality in Osasco of 0.7% (CI: 0.1; 1.3) in contrast to a mean annual decrease for Sorocaba of -1.5% (CI: -2.4; -0.6) and a stable average trend for São Paulo of -0.1 (IC: -0.3; 0.1). Similar increasing trends were seen in women. The age-period-cohort model showed an increase in the risk of death from 1996 in Osasco and a reduction for Sorocaba and São Paulo State during the same period. Our results point to a need for a special monitoring regarding lung cancer incidence and mortality in areas with higher asbestos exposure.

Lung cancer mortality in Australia: Projected outcomes to 2040

OBJECTIVES

The aim was to develop and validate a statistical model which uses past trends for lung cancer mortality and historical and current data on tobacco consumption to project lung cancer mortality rates into the future for Australia.

METHODS

We used generalized linear models (GLMs) with Poisson distribution including either age, birth cohort or period, and/or various measures of population tobacco exposure (considering cross-sectional smoking prevalence, cigarettes smoked and tar exposure per capita). Sex-specific models were fitted to data for 1956-2015 and age-standardized lung cancer mortality rates were projected forward to 2040. Possible lags of 20-30 years between tobacco exposure and lung cancer mortality were examined. The best model was selected using analysis of deviance. To validate the selected model, we temporarily re-fitted it to data for 1956-1990 and compared the projected rates to 2015 with the observed rates for 1991-2015.

RESULTS

The best fitting model used information on age, birth cohort and tar exposure per capita; close concordance with the observed data was achieved in the validation. The forward projections for lung cancer mortality using this model indicate that male and female age-standardized rates will decline over the period 2011-2015 to 2036-2040 from 27.2 to 15.1 per 100,000, and 15.8 to 11.8 per 100,000, respectively. However, due to population growth and ageing the number of deaths will increase by 7.9% for males and 57.9% for females; from 41,040 (24,831 males, 16,209 females) in 2011-2015 to 52,403 (26,805 males, 25,598 females) in 2036-2040.

CONCLUSION

In the context of the mature tobacco epidemic with past peaks in tobacco consumption for both males and females, lung cancer mortality rates are expected to continually decline over the next 25 years. However, the number of lung cancer deaths will continue to be substantial, and to increase, in Australia's ageing population.

A new age-period-cohort model for cancer surveillance research

We develop a new age-period-cohort model for cancer surveillance research; the theory and methods are broadly applicable. In the new model, cohort deviations are weighted to account for the variable number of periods that each cohort is observed. Weighting ensures that the fitted rates can be naturally expressed as a function of age × a function of period × a function of cohort. Furthermore, the age, period, and cohort deviations are split into orthogonal quadratic components plus higher-order terms. These decompositions enable powerful combination significance tests of first- and second-order age, period, and cohort effects. The regression parameters of the orthogonal quadratic polynomials (global curvatures) quantify how fast on average the trends in the rates are changing. Importantly, the global curvature for cohort determines the least squares slope of the expected annual percentage changes by age group versus age (local drifts), thereby providing a powerful one-degree-of-freedom test of age-period interactions. We introduce new estimable functions, including age gradients that quantify the rate of change of the longitudinal and cross-sectional age curves at each attained age, and gradient shifts that quantify how the cross-sectional age trend varies by period. We illustrate the new model using nationally representative multiple myeloma incidence. Comprehensive proofs are given in technical appendices. We provide an R package.

A unified approach for assessing heterogeneity in age-period-cohort model parameters using random effects

Age-period-cohort models are a popular tool for studying population-level rates; for example, trends in cancer incidence and mortality. Age-period-cohort models decompose observed trends into age effects that correlate with natural history, period effects that reveal factors impacting all ages simultaneously (e.g. innovations in screening), and birth cohort effects that reflect differential risk exposures that vary across birth years. Methodology for the analysis of multiple population strata (e.g. ethnicity, cancer registry) within the age-period-cohort framework has not been thoroughly investigated. Here, we outline a general model for characterizing differences in age-period-cohort model parameters for a potentially large number of strata. Our model incorporates stratum-specific random effects for the intercept, the longitudinal age trend, and the model-based estimate of annual percent change (net drift), thereby enabling a comprehensive analysis of heterogeneity. We also extend the standard model to include quadratic terms for age, period, and cohort, along with the corresponding random effects, which quantify possible stratum-specific departures from global curvature. We illustrate the utility of our model with an application to metastatic prostate cancer incidence (2004-2013) in non-Hispanic white and black men, using 17 population-based cancer registries in the Surveillance, Epidemiology, and End Results Program.