A divisive hierarchical clustering methodology for enhancing the ensemble prediction power in large scale population studies: the ATHLOS project

The ATHLOS cohort is composed of several harmonized datasets of international groups related to health and aging. As a result, the Healthy Aging index has been constructed based on a selection of variables from 16 individual studies. In this paper, we consider additional variables found in ATHLOS and investigate their utilization for predicting the Healthy Aging index. For this purpose, motivated by the volume and diversity of the dataset, we focus our attention upon data clustering, where unsupervised learning is utilized to enhance prediction power. Thus we show the predictive utility of exploiting hidden data structures. In addition, we demonstrate that imposed computation bottlenecks can be surpassed when using appropriate hierarchical clustering, within a clustering for ensemble classification scheme, while retaining prediction benefits. We propose a complete methodology that is evaluated against baseline methods and the original concept. The results are very encouraging suggesting further developments in this direction along with applications in tasks with similar characteristics. A straightforward open source implementation for the R project is also provided (https://github.com/Petros-Barmpas/HCEP).

Supplementary Information

The online version contains supplementary material available at 10.1007/s13755-022-00171-1.

COVID-19 and excess mortality in Russia: Regional estimates of life expectancy losses in 2020 and excess deaths in 2021

Accurately counting the human cost of the COVID-19 at both the national and regional level is a policy priority. The Russian Federation currently reports one of the higher COVID-19 mortality rates in the world; but estimates of mortality differ significantly. Using a statistical method accounting for changes in the population age structure, we present the first national and regional estimates of excess mortality for 2021; calculations of excess mortality by age, gender, and urban/rural status for 2020; and mean remaining years of life expectancy lost at the regional level. We estimate that there were 351,158 excess deaths in 2020 and 678,022 in 2021 in the Russian Federation; and, in 2020, around 2.0 years of life expectancy lost. While the Russian Federation exhibits very high levels of excess mortality compared to other countries, there is a wide degree of regional variation: in 2021, excess deaths expressed as a percentage of expected deaths at the regional level range from 27% to 52%. Life expectancy loss is generally greater for males; while excess mortality is greater in urban areas. For Russia as whole, an average person who died due to the pandemic in 2020 would have otherwise lived for a further 14 more years (and as high as 18 years in some regions), disproving the widely held view that excess mortality during the pandemic period was concentrated among those with few years of life remaining-especially for females. At a regional level, less densely populated, more remote regions, rural regions appear to have fared better regarding excess mortality and life expectancy loss-however, a part of this differential could be owing to measurement issues. The calculations demonstrate more clearly the true degree of the human cost of the pandemic in the Russian Federation.

Development of a common scale for measuring healthy ageing across the world: results from the ATHLOS consortium

BACKGROUND

Research efforts to measure the concept of healthy ageing have been diverse and limited to specific populations. This diversity limits the potential to compare healthy ageing across countries and/or populations. In this study, we developed a novel measurement scale of healthy ageing using worldwide cohorts.

METHODS

In the Ageing Trajectories of Health-Longitudinal Opportunities and Synergies (ATHLOS) project, data from 16 international cohorts were harmonized. Using ATHLOS data, an item response theory (IRT) model was used to develop a scale with 41 items related to health and functioning. Measurement heterogeneity due to intra-dataset specificities was detected, applying differential item functioning via a logistic regression framework. The model accounted for specificities in model parameters by introducing cohort-specific parameters that rescaled scores to the main scale, using an equating procedure. Final scores were estimated for all individuals and converted to T-scores with a mean of 50 and a standard deviation of 10.

RESULTS

A common scale was created for 343 915 individuals above 18 years of age from 16 studies. The scale showed solid evidence of concurrent validity regarding various sociodemographic, life and health factors, and convergent validity with healthy life expectancy (r = 0.81) and gross domestic product (r = 0.58). Survival curves showed that the scale could also be predictive of mortality.

CONCLUSIONS

The ATHLOS scale, due to its reliability and global representativeness, has the potential to contribute to worldwide research on healthy ageing.

Alcohol Drinking and Health in Ageing: A Global Scale Analysis of Older Individual Data through the Harmonised Dataset of ATHLOS

We investigated the relation between alcohol drinking and healthy ageing by means of a validated health status metric, using individual data from the Ageing Trajectories of Health: Longitudinal Opportunities and Synergies (ATHLOS) project. For the purposes of this study, the ATHLOS harmonised dataset, which includes information from individuals aged 65+ in 38 countries, was analysed (n = 135,440). Alcohol drinking was reflected by means of three harmonised variables: alcohol drinking frequency, current and past alcohol drinker. A set of 41 self-reported health items and measured tests were used to generate a specific health metric. In the harmonised dataset, the prevalence of current drinking was 47.5% while of past drinking was 26.5%. In the pooled sample, current alcohol drinking was positively associated with better health status among older adults ((b-coef (95% CI): 1.32(0.45 to 2.19)) and past alcohol drinking was inversely related (b-coef (95% CI): -0.83 (-1.51 to -0.16)) with health status. Often alcohol consumption appeared to be beneficial only for females in all super-regions except Africa, both age group categories (65-80 years old and 80+), both age group categories, as well as among all the financial status categories (all p < 0.05). Regional analysis pictured diverse patterns in the association for current and past alcohol drinkers. Our results report the need for specific alcohol intake recommendations among older adults that will help them maintain a better health status throughout the ageing process.

AN INTRODUCTION TO THE ATHLOS PROJECT: AGEING TRAJECTORIES OF HEALTH: LONGITUDINAL OPPORTUNITIES AND SYNERGIES

ATHLOS is a 5-year project, funded by the European Union’s Horizon 2020 Research and Innovation Program. Its aim is to achieve a better understanding of healthy ageing, utilising longitudinal data from existing cohort studies. The measure of healthy ageing used within ATHLOS is based on the definition used by the World Health Organization as the ongoing process of developing and maintaining functional ability to enable well-being in older age. The first step of the project was to harmonise 17 community based cohort studies of ageing, covering 38 countries over the world and over 411,000 individuals. In this talk we will discuss the work of the different work packages of the project, including a description of the existing evidence on risk factors of healthy ageing.

Is the demographic dividend an education dividend?

The effect of changes in age structure on economic growth has been widely studied in the demography and population economics literature. The beneficial effect of changes in age structure after a decrease in fertility has become known as the "demographic dividend." In this article, we reassess the empirical evidence on the associations among economic growth, changes in age structure, labor force participation, and educational attainment. Using a global panel of countries, we find that after the effect of human capital dynamics is controlled for, no evidence exists that changes in age structure affect labor productivity. Our results imply that improvements in educational attainment are the key to explaining productivity and income growth and that a substantial portion of the demographic dividend is an education dividend.

Effects on well-being of investing in cleaner air in India

Over the past decade, India has experienced rapid economic growth along with increases in levels of air pollution. Our goal is to examine how alternative policies for air pollution abatement affect well-being there. In particular, we estimate the effects of policies to reduce the levels of ambient fine particulates (PM2.5), which are especially harmful to human health, on well-being, quantified using the United Nations' human development index (HDI). Two of the three dimensions of this index are based on gross domestic product (GDP) per capita and life expectancy. Our approach allows reductions in PM2.5 to affect both of them. In particular, economic growth is affected negatively through the costs of the additional pollution control measures and positively through the increased productivity of the population. We consider three scenarios of PM2.5 abatement, corresponding to no further control, current Indian legislation, and current European legislation. The overall effect in both control scenarios is that growth in GDP is virtually unaffected relative to the case of no further controls, life expectancy is higher, and well-being, as measured by the HDI, is improved. In India, air pollution abatement investments clearly improve well-being.

Pesticide use and practices in an Iowa farm family pesticide exposure study

Residents of Iowa were enrolled in a study investigating differences in pesticide contamination and exposure factors between 25 farm homes and 25 non-farm homes. The target pesticides investigated were atrazine, metolachlor, acetochlor, alachlor, 2,4-D, glyphosate, and chlorpyrifos; all were applied to either corn or soybean crops. A questionnaire was administered to all participants to determine residential pesticide use in and around the home. In addition, a questionnaire was administered to the farmers to determine the agricultural pesticides they used on the farm and their application practices. Non-agricultural pesticides were used more in and around farm homes than non-farm homes. Atrazine was the agricultural pesticide used most by farmers. Most farmers applied pesticides themselves but only 10 (59%) used tractors with enclosed cabs, and they typically wore little personal protective equipment (PPE). On almost every farm, more than one agricultural pesticide was applied. Corn was grown by 23 (92%) farmers and soybeans by 12 (48%) farmers. Of these, 10 (40%) grew both soybeans and corn, with only 2 (8%) growing only soybeans and 13 (52%) growing only corn. The majority of farmers changed from their work clothes and shoes in the home, and when they changed outside or in the garage, they usually brought their clothes and shoes inside. Applying pesticides using tractors with open cabs, not wearing PPE, and changing from work clothes in the home may increase pesticide exposure and contamination. Almost half of the 66 farm children less than 16 years of age were engaged in some form of farm chores, with 6 (9%) potentially directly exposed to pesticides, while only 2 (4%) of the 52 non-farm children less than 16 years of age had farm chores, and none were directly exposed to pesticides. Farm homes may be contaminated with pesticides in several ways, resulting in potentially more contamination than non-farm homes, and farm children may be directly exposed to pesticides through farm chores involving pesticides. In addition to providing a description of pesticide use, the data presented here will be useful in evaluating potential contributing factors to household pesticide contamination and family exposure.

The end of world population growth

There has been enormous concern about the consequences of human population growth for the environment and for social and economic development. But this growth is likely to come to an end in the foreseeable future. Improving on earlier methods of probabilistic forecasting, here we show that there is around an 85 per cent chance that the world's population will stop growing before the end of the century. There is a 60 per cent probability that the world's population will not exceed 10 billion people before 2100, and around a 15 per cent probability that the world's population at the end of the century will be lower than it is today. For different regions, the date and size of the peak population will vary considerably.

Kidney disease and arthritis in a cohort study of workers exposed to silica

Silica exposure has been associated with kidney disease and rheumatoid arthritis; an autoimmune mechanism has been proposed. Approximately 2 million people are occupationally exposed to silica in the United States, 100,000 at more than twice the National Institute for Occupational Safety and Health recommended exposure limit of 0.05 mg/m(3). We examined renal disease morbidity and mortality, as well as arthritis mortality, in a cohort of 4,626 silica-exposed workers in the industrial sand industry (an industry previously unstudied). We compared the cohort with the U.S. population and also conducted internal exposure-response analyses using a job-exposure matrix based on more than 4,000 industrial hygiene samples. We found excess mortality from acute renal disease [standardized mortality ratio (SMR) = 2.61, 95% confidence intervals (95% CIs) = 1.49--4.24; 16 deaths], chronic renal disease (SMR = 1.61, 95% CI = 1.13--2.22; 36 deaths), and arthritis (SMR = 4.36, 95% CI = 2.76--6.54; 23 deaths) on the basis of multiple-cause mortality data, which considered any mention of disease on a death certificate. Linking the cohort with the U.S. registry of end-stage renal disease for the years 1977-1996, we found an excess of end-stage renal disease incidence (standardized incidence ratio = 1.97, 95% CI = 1.25--2.96; 23 cases), which was highest for glomerulonephritis (standardized incidence ratio = 3.85, 95% CI = 1.55--7.93; 7 cases). We found increasing end-stage renal disease incidence with increasing cumulative exposure; standardized rate ratios by quartile of cumulative exposure were 1.00, 3.09, 5.22, and 7.79. A positive exposure-response trend was also observed for rheumatoid arthritis on the basis of death certificate data. These data represent the largest number of kidney disease cases analyzed to date in a cohort with well-defined silica exposure and suggest a causal link between silica and kidney disease. Excess risk of end-stage renal disease due to a lifetime of occupational exposure at currently recommended limits is estimated to be 14%, above a background end-stage renal disease risk of 2%.

Lung cancer among industrial sand workers exposed to crystalline silica

In 1997, the International Agency for Research on Cancer determined that crystalline silica was a human carcinogen but noted inconsistencies in the epidemiology. There are few exposure-response analyses. The authors examined lung cancer mortality among 4,626 industrial sand workers, estimating exposure via a job-exposure matrix based on 4,269 industrial hygiene samples collected in 1974--1995. The average length of employment was 9 years, and estimated average exposure was 0.05 mg/m(3) (the National Institute of Occupational Safety and Health Recommended Exposure Limit). Results confirmed excess mortality from silicosis/pneumoconioses (standardized mortality ratio = 18.2, 95% confidence interval: 10.6, 29.1; 17 deaths). The lung cancer standardized mortality ratio was 1.60 (95% confidence interval: 1.31, 1.93; 109 deaths). Limited data suggested that smoking might account for 10--20% of the lung cancer excess. Exposure-response analyses by quartile of cumulative exposure (15-year lag) yielded standardized rate ratios of 1.00, 0.78, 1.51, and 1.57 (p for trend = 0.07). Nested case-control analyses after exclusion of short-term workers, who had high overall morality, yielded odds ratios by quartile of cumulative exposure (15-year lag) of 1.00, 1.35, 1.63, and 2.00 (p for trend = 0.08) and odds ratios by quartile of average exposure of 1.00, 0.92, 1.44, and 2.26 (p = 0.005). These data lend support to the labeling by the International Agency for Research on Cancer of silica as a human carcinogen. There are approximately 2 million US workers exposed to silica; 100,000 are exposed to more than 0.1 mg/m(3).

Impact of a worker notification program: assessment of attitudinal and behavioral outcomes

BACKGROUND

Hundreds of worker notification programs are conducted each year to communicate occupational risks to workers. However, few attempts have been made to evaluate their effectiveness and few have described how communication theories are applied in developing notification messages. We developed and assessed the effectiveness of a worker notification program at a beryllium machining plant.

METHODS

We compared self-protective attitudinal and behavioral responses among workers in two plants: (1) an intervention plant that received beryllium risk notification and (2) a matched control plant that did not receive notification.

RESULTS

Workers receiving notification reported significantly stronger perceptions of threat and efficacy, more positive attitudes toward safety practices, and engaged in more protective behaviors than the workers at the control plant.

CONCLUSIONS

This study demonstrates the utility of applying communication theories in the development of notification messages and the results suggest that mass presentations may be just as effective, if not more so, than one-on-one notifications. Am. J. Ind. Med. 37:205-213, 2000. Published 2000 Wiley-Liss, Inc.

Academic psychiatry and managed care: a case study

An academic department of psychiatry in New York City eliminated the need for behavioral managed care intermediaries by transforming itself from a fee-for-service system to a system able to engage in full-risk capitation contracts. The first step was to require health maintenance organizations to contract directly with the department. The department formed two legal entities, a behavioral management services organization for utilization management and a behavioral integrated provider association. The authors describe these entities and review the first year of operation, presenting data on enrollees, capitation rates, and service utilization for the first three contracts. The fundamental differences in the treatment model under managed care and under a fee-for-service system are highlighted. The authors conclude that by contracting directly with insurers on a full-risk capitation basis, departments of psychiatry will be better able to face the economic threats posed by the cost constraints inherent in managed care and maintain or re-establish their autonomy as care managers as well as high-quality care providers.

Doubling of world population unlikely

Most national and international agencies producing population projections avoid addressing explicitly the issue of uncertainty. Typically, they provide either a single projection or a set of low, medium and high variants, and only very rarely do they give these projections a probabilistic interpretation. Probabilistic population projections have been developed for specific industrialized countries, mostly the United States, and are based largely on time-series analysis. On a global level, time-series analysis is not applicable because there is a lack of appropriate data, and for conceptual reasons such as the structural discontinuity caused by the demographic transition. Here we report on a new probabilistic approach that makes use of expert opinion on trends in fertility, mortality and migration, and on the 90 per cent uncertainty range of those trends in different parts of the world. We have used simulation techniques to derive probability distributions of population sizes and age structures for 13 regions of the world up to the year 2100. Among other things, we find that there is a probability of two-thirds that the world's population will not double in the twenty-first century.

Low frequency oscillations in pulmonary arterial pressure in chronic heart failure

Spontaneous, sustained low frequency oscillations of pulmonary arterial pressure were observed in 8 patients with moderate to severe chronic heart failure during recordings of pulmonary arterial pressure of 8.6 to 48 hours duration. The oscillations (frequency range 0.015 to 0.035 Hz) had a distinct peak from and lower frequency range than respiration (0.2 to 0.6 Hz) and heart rate (0.8 to 1.9 Hz). They were present from 42% to 82% of the recording time with an amplitude varying from undetectable to a maximum range from 5.3 mm Hg to 19.7 mm Hg. When detectable, the mean amplitude of the low frequency oscillations of pulmonary arterial pressure ranged from 1.2 +/- 1.2(SD) mm Hg to 4.3 +/- 3.3 mm Hg. These oscillations are lower in frequency than the low frequency component usually described in recordings of systemic arterial pressure and heart rate in normal subjects, and are closer to the very low frequency rhythms described in severe chronic heart failure and Cheyne Stokes respiration.

Outbreak of hypersensitivity pneumonitis among mushroom farm workers

Between April 1982 and August 1985, seven cases of mushroom worker's lung (MWL), a form of hypersensitivity pneumonitis, were diagnosed among workers at one mushroom farm in Florida. The cases suffered from episodic shortness of breath, cough, fever and chills, myalgia, malaise, and difficulty breathing. Pulmonary function testing revealed restrictive ventilatory impairment and reduced diffusing capacity; chest radiographs exhibited diffuse interstitial pulmonary infiltrates. The seven cases occurred among workers from different farm operations, suggesting that workers throughout the farm were exposed to the disease causing agent(s). Six of the affected workers left employment at the farm in order to remain free of symptoms. The other affected worker was able to continue working at the farm, but only by remaining in a maintenance shop which was physically separated from the rest of the farm facilities. An industrial hygiene survey demonstrated that farm workers from every work area were exposed to organic dust constituents suspected of causing MWL, but no specific antigens were identified as the cause of the cases. Of the remaining workers who participated in a cross-sectional respiratory morbidity survey at the farm, approximately 20% of the more heavily exposed workers reported occasionally experiencing symptoms consistent with MWL. Approximately 10% of the workers had below normal spirometry test results, but interpretation was hampered by the diverse racial makeup of the population and lack of an adequate comparison group. No abnormalities consistent with either acute or chronic MWL were seen on the chest radiographs. Serologic tests demonstrated that almost all workers had been exposed to antigens capable of causing MWL, but the results were not associated with health status. At the time of the cross-sectional survey, no workers were found to be suffering acute respiratory problems consistent with MWL.