BackgroundLoneliness is a public health priority, yet population-scale evidence that distinguishes types of loneliness and links with living alone circumstances is scarce. Population-wide linked administrative data within the Secure Anonymised Information Linkage (SAIL) Databank provide an opportunity to investigate these relationships across Wales. Objectives(1) Describe long-term living alone trends; (2) develop indicators for reason for living alone; and (3) examine associations between living alone (including derived indicators) and emotional, social and overall loneliness, alongside demographic, health and place-based factors. MethodsWe conducted a two-phase observational study in Wales. Phase 1 used annual, population level snapshots (2008--2024) to quantify living alone trends and correlates. Phase 2 linked individual level data in SAIL to National Survey for Wales loneliness data (De Jong Gierveld 6-item) to model adjusted associations. We inferred living alone reasons (partner loss, separation, unassigned) from life events data. Models adjusted for age and gender interaction, health, tenure, housing type, deprivation, and built environment indicators. ResultsLiving alone households increased from 27.0% (314,806) in 2008 to 29.6% (363,133) in 2024. In adjusted models, living alone was strongly associated with loneliness risk, especially emotional loneliness. These associations were modified by reason for living alone, with the largest effects observed following recent partner loss. Younger adults reported highest overall loneliness; men showed higher social loneliness around midlife; women showed higher emotional loneliness. Built environment measures showed little evidence of independent association after adjustment. ConclusionsPopulation-scale linkage can generate robust, policy-relevant indicators of living alone dynamics. Such indicators support monitoring and targeting within health and social care, complementing national statistics and enabling future linked data studies.
Infant mortality is an important indicator of population health and development outcomes. This study examines the associations between age at first marriage, poverty, total fertility rate (TFR), and provincial level infant mortality in Indonesia. Using an ecological cross-sectional design based on provincial level secondary data from 35 provinces, this study applies multiple linear regression to examine how demographic and socioeconomic indicators are associated with variations in infant mortality rate (IMR) across provinces. The regression model explained 77.1% of the variation in provincial IMR (R2=0.771). Poverty and total fertility rate showed positive and statistically significant associations with infant mortality, indicating that provinces with higher poverty rates and higher fertility levels tended to report higher IMR. In contrast, age at first marriage showed a negative but statistically non-significant association with IMR after controlling for poverty and fertility. These findings indicate that socioeconomic and fertility-related conditions showed stronger statistical associations with provincial differences in infant mortality than age at first marriage within the present model. However, the findings should be interpreted cautiously because the study uses aggregated provincial level data and an ecological cross-sectional design, which does not permit causal or individual-level inference.
IntroductionIn the context of chronic pain, no study has yet compared risk-based and count-based approaches for modelling medication use trajectories. Risk-based approach reflects the potential burden of medication-related adverse effects, whereas count-based approach reflects treatment burden and regimen complexity. Comparing these complementary measures may identify different patient profiles depending on the clinical or research objective. ObjectivesThis study aimed to compare a composite medication risk score, the Medication Quantification Scale 4.0 (MQS-4.0), with medication count in examining medication use trajectories among individuals with chronic pain. MethodsWe conducted a cohort study using data from the TorSaDE Cohort, which links five cycles of Statistics Canada's Canadian Community Health Survey (2007-2016 cross-sectional questionnaires) and Québec provincial health administrative databases (1996-2016 longitudinal data). We selected 8,760 adults with chronic pain who were covered by public prescription drug insurance. Pain-related medications used in the two-year period following survey completion were operationalised as monthly risk scores and monthly medication count. Growth Mixture Modelling was applied to each of these measures to identify groups with similar patterns over time (trajectories). ResultsBoth the MQS-4.0 risk score and the medication count were effective in forming trajectory groups that were statistically robust (based on entropy and Bayesian information criterion) and clinically meaningful. However, the MQS-4.0 risk score approach was more discriminative, identifying differences between trajectory groups in terms of sociodemographic characteristics, pain profiles, health status, and healthcare utilisation. Individuals with the highest pain-related medication count did not necessarily have the highest MQS-4.0 risk scores, and some individuals using fewer medications were found to be at high risk. ConclusionsOur findings suggest that a risk-based approach such as the MQS-4.0 may provide complementary information to a count-based approach for subgroup analyses, particularly when using health administrative databases to study pain-related medication use in individuals living with chronic pain.
IntroductionDisabled people experienced a disproportionate burden of COVID-19, yet evidence on how household living arrangements shaped these outcomes across disability groups remains limited. ObjectiveTo describe the creation, content, and baseline characteristics of a population-based linked administrative cohort for studying disability, residential context, and COVID-19 infection, hospitalisation, and mortality in Scotland. MethodsThe cohort was assembled from Community Health Index records for people alive and resident in Scotland on 1 March 2020 whose residential addresses could be linked to Unique Property Reference Numbers. Records for those aged under 16 years at that date were excluded, and the remainder were linked to the 2011 Scottish Census to obtain disability status and sociodemographic characteristics. Disability was classified sequentially into three mutually exclusive groups: intellectual disability, including individuals who may also have reported physical or sensory disability; physical disability, including sensory impairments but excluding intellectual disability; and a comparison group without reported intellectual or physical disability. Household context was characterised by residential setting and household size. Residential setting was classified from Basic Land and Property Unit codes as private dwellings, small communal establishments (fewer than 50 linked residents), or large communal establishments (50 or more linked residents). Household size was grouped as 1, 2--4, 5--7, 8--10, or 11 or more linked residents. Laboratory-confirmed SARS-CoV-2 infection and COVID-19-related hospitalisation and mortality were ascertained from national surveillance and administrative records between 1 March 2020 and 30 April 2022. Age- and sex-adjusted incidence rates were estimated using Poisson regression with log person-time as an offset. ResultsOf 4,132,960 age-eligible Community Health Index records, 3,719,651 adults (90.0%) were successfully linked to the Census and included in the cohort. Of these, 17,354 had intellectual disability, 377,706 had physical disability and 3,324,591 were in the comparison group. During follow-up, 306,343 cohort members had a laboratory-confirmed SARS-CoV-2 infection, 22,945 were hospitalised with COVID-19 and 12,893 died from COVID-19-related causes. The corresponding crude incidence rates were 592 infections, 32 hospitalisations and 16 deaths per 10,000 person-years. Age- and sex-adjusted rates were highest among adults with intellectual disability, at 1,486 infections, 86 hospitalisations and 56 deaths per 10,000 person-years. The corresponding rates were 932, 58 and 33 among adults with physical disability, and 570, 29 and 12 in the comparison group. Across all three outcomes, adjusted incidence rates were higher in communal establishments than in private dwellings and generally increased with household size. Mortality differences were especially pronounced in communal establishments and at addresses with 11 or more linked residents. ConclusionsThis population-based, Census-linked cohort integrates information on disability and residential context with national health and administrative records. Although developed to examine COVID-19 outcomes, it provides a broader platform for investigating how disability and living arrangements shape health inequalities in Scotland.
IntroductionDespite the established evidence that children in foster homes will face a high burden of health challenges, data on their long-term health is often not accessible due to a lack of centralised data systems, cost of longitudinal studies, and lost to follow-up. We present the CANadian children in FOSter homes (CanFos) cohorts with linked longitudinal health data to enable population-level studies on the health of children in foster homes across Canada. MethodsCanFos holds multiple national Census-based cohorts at 5-year cycles on children in foster homes, along with all other Canadian children as a comparison population, linked to health administrative data to enable ongoing follow-up of outcomes. Socio-demographic and economic information is available at the individual and household level. Data linkage is renewed routinely with data currently available from 2000 until 2024. ResultsTo date, linked data includes weighted N= 31 595, N= 24 125, N=22 595 children less than 18 years of age living in foster homes in 2011, 2016, and 2021 compared to over 5 million other Canadian children. Rates of living in private dwelling foster homes was 5.9 per 1 000 children (2011); 4.4 per 1 000 children (2016); 4.0 per 1 000 children (2021). For those who lived in foster homes, at twelve years of follow-up the number of hospitalisations was 345 per 1 000 persons and the crude counts of mortality was 22 per 1 000 (vs 211 hospitalisations per 1 000 persons, and 4 deaths per 1 000 persons for other Canadians). ConclusionCanFos is a comprehensive data linkage resource that can ask multiple research questions about the patterns of health care use, health outcomes, and mortality of children in foster homes. Ultimately, the evidence has the potential to lead to policies and health programs that safeguard the health of children in foster care.
