Background: The counterfactual definition of confounding is often explained in the context of exchangeability between the exposed and unexposed groups. One recent approach is to examine whether the measures of association (eg, associational risk difference) are exchangeable when exposure status is flipped in the population of interest. We discuss the meaning and utility of this approach, showing their relationships with the concept of confounding in the counterfactual framework.
Methods: Three hypothetical cohort studies are used, in which the target population is the total population. After providing an overview of the notions of confounding in distribution and in measure, we discuss the approach from the perspective of exchangeability of measures of association (eg, factual associational risk difference vs counterfactual associational risk difference).
Results: In general, if the measures of association are non-exchangeable when exposure status is flipped, confounding in distribution is always present, although confounding in measure may or may not be present. Even if the measures of association are exchangeable when exposure status is flipped, there could be confounding both in distribution and in measure. When we use risk difference or risk ratio as a measure of interest and the exposure prevalence in the population is 0.5, testing the exchangeability of measures of association is equivalent to testing the absence of confounding in the corresponding measures.
Conclusion: The approach based on exchangeability of measures of association essentially does not provide a definition of confounding in the counterfactual framework. Subtly differing notions of confounding should be distinguished carefully.
Background: Serial weight decrease can be a prognostic predictor in chronic hemodialysis (HD) patients. We investigated the impact of long-term post-HD body weight (BW) changes on all-cause mortality among HD patients.
Methods: This longitudinal cohort study and post-hoc analysis evaluated participants of a previous randomized controlled trial conducted between 2006 and 2011 who were followed up until 2018. Weight change slopes were generated with repeated measurements every 6 months during the trial for patients having ≥5 BW measurements. Participants were categorized into four groups based on quartiles of weight change slopes; the median weight changes per 6 months were −1.02 kg, −0.25 kg, +0.26 kg, and +0.86 kg for first, second, third, and fourth quartile, respectively. Cox proportional hazard regression was used to evaluate differences in subsequent survival among the four groups. BW trajectories were plotted with a backward time-scale and multilevel regression analysis to visualize the difference in BW trajectories between survivors and non-survivors.
Results: Among the 461 patients, 404 were evaluated, and 168 (41.6%) died within a median follow-up period of 10.2 years. The Cox proportional hazard regression adjusted for covariates and baseline BW showed that a higher rate of weight loss was associated with higher mortality. The hazard ratios were 2.02 (95% confidence interval [CI], 1.28–3.20), 1.77 (95% CI, 1.10–2.85), 1.00 (reference), and 1.11 (95% CI, 0.67–1.83) for the first, second, third (reference), and fourth quartiles, respectively. BW trajectories revealed a significant decrease in BW in non-survivors.
Conclusion: Weight loss elucidated via serial BW measurements every 6 months is significantly associated with higher mortality among HD patients.
Background: Cold exposure induces lower urinary tract symptoms, including nocturia. Cold-induced detrusor overactivity can be alleviated by increasing skin temperature in rats. However, no study has shown an association between passive heating via hot-water bathing and nocturia among humans.
Methods: We included 1,051 Japanese community-dwelling older adults (mean age: 71.7 years) in this cross-sectional study from 2010 to 2014. The number of nocturnal voids was recorded in a self-administered urination diary. Nocturia was defined as ≥2 nocturnal voids. We evaluated bathing conditions in the participants’ houses.
Results: Hot-water bathing (n = 888) was associated with a lower prevalence of nocturia than no bathing (n = 163), independent of potential confounders, including age, sex, obesity, income, physical activity, diabetes, medication (diuretics, nondiuretic antihypertensives, and hypnotics), depressive symptoms, indoor/outdoor temperature, and day length (odds ratio [OR] 0.68; 95% confidence interval [CI], 0.48–0.97; P = 0.035). Compared with the quartile group with the longest bath-to-bed interval (range: 161–576 min), the second and third quartile groups (range: 61–100 and 101–160 min, respectively) were associated with a lower prevalence of nocturia, after adjusting for water temperature and bathing duration besides the same covariates (OR 0.60; 95% CI, 0.38–0.96; P = 0.031 and OR 0.59; 95% CI, 0.37–0.94; P = 0.025, respectively).
Conclusion: Hot-water bathing, particularly with a bath-to-bed interval of 61–160 min, was significantly associated with a lower prevalence of nocturia among older adults.
Background: Identifying which exposures cause disease and quantifying their impacts is essential in promoting and monitoring public health. When multiple exposures are involved, measuring individual contributions becomes challenging.
Methods: The authors propose a disease attribution method based on aggregate data or summary statistics of individual-level data, possibly from multiple data sources.
Results: Using the proposed method, the burden of disease is apportioned to the independent and interaction effects of each of its major risk factors and all the other factors as a whole. This scheme guarantees that 100% is the total share of the burden.
Conclusion: The calculation is simple and straightforward; therefore, it is recommended for use in studies on disease burden.
Background: It has not been determined whether mentally active sedentary behavior (MASB) and passive sedentary behavior (PSB) differentially affect cognitive function and whether these associations differ according to physical activity (PA) level. We examined the comparative impacts of MASB and PSB on dementia onset and aimed to understand whether the associations differed by PA level.
Methods: We conducted a 5-year longitudinal study involving all community-dwelling older adults in a rural area in Japan (n = 5,323). Dementia onset was examined using long-term care insurance data. PA was evaluated using the International Physical Activity Questionnaire and categorized as low (<2.5 metabolic equivalent of task [MET]-h/week), moderate (2.5–16.0 MET-h/week), or high (≥16.0 MET-h/week). We also assessed PSB (TV-watching time; <1 h/day, 1–3 h/day, ≥3 h/day) and MASB (Book-reading time; <10 min/day, 10–30 min/day, ≥30 min/day). To examine the associations of MASB and PSB with dementia onset, we performed the Fine-Gray models accounting for competing risk of death.
Results: During the follow-up period, 606 (11.4%) participants developed dementia. MASB was independently associated with a lower risk of dementia; the magnitude of the impact was significant at higher PA levels. There was no association between PSB and developing dementia across all PA levels. Furthermore, dementia risk for individuals with high PA levels and moderate or high MASB levels was approximately 60% lower than those with low PA levels and low MASB.
Conclusion: Providing interventions to promote MASB, which reduces dementia risk, and PA, which increases MASB’s effect on dementia incidence, can be beneficial in delaying or preventing dementia onset.
Background: Recent innovations in information and communication technology have made it possible to assess diet using web-based methods; however, their applicability in the general population remains unclear. Hence, we aimed to examine the applicability of a web-based 24-hour dietary recall (24HR) tool to large-scale epidemiological studies by determining the sampling rate and characteristics of randomly selected participants from a Japanese cohort study.
Methods: In total, 5,013 individuals were recruited from a cohort of 21,537 individuals, and 975 agreed to participate in this study. The participants selected either self-administered web-based dietary 24HR (self-administered 24HR) or interviewer-administered telephone-based 24HR (interviewer-administered 24HR) as the method for the dietary assessment and answered questions regarding the acceptability of the system.
Results: The response rate of the 975 participants was 19.4%, corresponding to approximately 4.5% of the total study sample. About half of them chose the self-administered 24HR (46.9%). The median time required for the self-administered and interviewer-administered 24HR was 25 and 27 minutes, respectively. In the self-administered 24HR, older people, regardless of sex, tended to require a longer time, and approximately 60% of the participants rated the ease of use of the system as “somewhat difficult” or “difficult.”
Conclusion: Characteristics of the participants in this study were not systemically different from those of the entire study sample. Improvements in the approach to entering cooking details and the dish name selection may be necessary for better acceptability in order to be accepted as a self-administered dietary recall tool.
Background: The Longevity Improvement & Fair Evidence (LIFE) Study, which was launched in 2019, is a multi-region community-based database project that aims to generate evidence toward extending healthy life expectancy and reducing health disparities in Japan. Herein, we describe the LIFE Study’s design and baseline participant profile.
Methods: Municipalities participating in the LIFE Study provide data from government-administered health insurance enrollees and public assistance recipients. These participants cover all disease types and age groups. Centered on healthcare claims data, the project also collects long-term care claims data, health checkup data, vaccination records, residence-related information, and income-related information. The different data types are converted into a common data model containing five modules (health care, long-term care, health checkup, socioeconomic status, and health services). We calculated the descriptive statistics of participants at baseline in 2018.
Results: The LIFE Study currently stores data from 1,420,437 residents of 18 municipalities. The health care module contains 1,280,756 participants (mean age: 65.2 years), the long-term care module contains 189,069 participants (mean age: 84.3 years), and the health checkup module contains 274,375 participants (mean age: 69.0 years). Although coverage and follow-up rates were lower among younger persons, the health care module includes 74,151 children (0–19 years), 273,157 working-age adults (20–59 years), and 933,448 older persons (≥60 years).
Conclusion: The LIFE Study provides data from over 1 million participants and can facilitate a wide variety of life-course research and cohort studies. This project is expected to be a useful platform for generating real-world evidence from Japan.