Search

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).

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.

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.

Early life growth has been suggested to influence bone health. However, the relationship with risk of hip fracture in old age has not been thoroughly investigated. We therefore studied the association between birth weight and hip fracture incidence after age 50 among 10 893 men and women (48% women) from the Uppsala Birth Cohort Study (UBCoS, born 1915–1929) and 1334 men from the Uppsala Longitudinal Study of Adult Men (ULSAM, born 1920–1924). Birth weight was collected from hospital or midwives’ records and hip fractures were obtained from the Swedish Hospital Discharge Register. We observed 717 hip fractures in UBCoS (458 in women, 259 in men, end of follow-up: 31 December 2008) and 102 hip fractures in ULSAM (end of follow-up: 31 December 2009). There were no indications of non-linear associations. Results are presented as hazard ratios (HR) and 95% CI per 1 kg increase in birth weight. The crude HR for 1 kg increase in birth weight on hip fracture rate in UBCoS was 0.99 (95% CI: 0.85–1.14). After controlling for gender and socioeconomic status at birth, the HR was 1.06 (95% CI: 0.91–1.23). Additional adjustment for adult height and comorbidity in a subgroup of UBCoS men (n=1241, 50 hip fractures) gave a HR of 0.97 (95% CI: 0.52–1.80). Parity and gestational age did not largely influence the estimates. Neither birth weight standardized for gestational age nor gestational duration was associated with hip fracture rate. The unadjusted HR in ULSAM was 1.06 (95% CI: 0.73–1.53). After adjustment for adult body mass index, height, social class, comorbidity, and smoking status, the HR was 1.03 (95% CI: 0.70–1.51). Based on the results from two population-based cohorts with accurate assessment of both birth weight and hip fractures, we conclude that there is no association between birth weight and risk of hip fracture.

Objective: To study whether the effect of size at birth on the risk of ischaemic heart disease (IHD) death is modified by social circumstances in childhood or in adulthood. Design: A cohort study. Data on circumstances at birth were retrieved from archived obstetric records, social characteristics in adulthood and mortality follow-up through routine registers. Participants: 6159 men and 5663 women who were born in Uppsala University Hospital, Sweden (the Uppsala Birth Cohort) during 1915–1929, were singleton births with more than 30 weeks of gestational age and were alive in 1961. Follow-up time 1961–2002 (from age 31–46 to 73–88 years). Main outcome measure: Death from IHD. Multivariate Cox regression with age as the time scale, controlling for year of birth and stratified by gender. Results: The risk of IHD death was lower among men and women with higher weight for gestational age. Lower social class in adulthood was associated with a higher risk of IHD death. The effect of size at birth on IHD mortality did not appear to be modified by social class at birth but was only present in men of higher social class in adulthood (hazard ratio per 1 SD weight for gestational age 0.84, 95% CI 0.75 to 0.93). Conclusions: Weight for gestational age was inversely associated with the risk of IHD death in men and women; this effect was present in men of non-manual adult social class only but did not appear to be modified by adult social class in women or by social class at birth in either men or women.

Objectives:To investigate the combined effects of size at birth and maternal education on prevalence of overweight and obesity among 18-year-old men.Methods:We studied the associations of weight for gestational age and maternal education with body mass index (BMI), overweight and obesity by multivariate linear and logistic regression, adjusted for mother's age, parity and diabetes in a register-based cohort of 6535 men born between 1973 and 1985. Further adjustments for mother's height, pre-pregnancy weight, weight gain and smoking during pregnancy were made in a subsample of 1103 men born between 1982 and 1985.Results:Mean BMI and prevalence of overweight and obesity decreased with higher maternal education. Mother's BMI and smoking were the strongest predictors of sons' overweight and obesity and essentially accounted for the variation in son's overweight by maternal education. The association of size at birth with later overweight was present only in sons born to mothers who were nonsmokers (odds ratio per 1 standard deviation weight for gestational age z-score 1.46, 95% CI 1.18-1.81) and became substantially reduced on adjustment for mother's pre-pregnancy BMI. Length of gestation was not statistically significantly associated with BMI at age 18.Conclusions:Maternal overweight and maternal smoking were the strongest determinants of offspring overweight and its social patterning, and should be a priority for public health policies.International Journal of Obesity (2008) 32, 73-81; doi:10.1038/sj.ijo.0803681; published online 31 July 2007.