PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 10474137

Dispersion models and longitudinal data analysis.

B Jorgensen·M Tsao
View original source ↗Export metadata ↓

Abstract

Dispersion models provide a flexible class of non-normal distributions with many potential applications in biostatistics, accommodating a wide range of continuous, discrete and mixed data. Starting with Liang and Zeger's generalized estimating equation method, we review some recent applications of dispersion models in longitudinal data analysis, including state space models based on the Tweedie class of exponential dispersion models. In medical applications the latent process of a state space model may often be interpreted as an unobserved potential morbidity process, which is modelled as a function of time varying covariates. By allowing a multivariate response vector of 'symptoms', the model integrates several response variables of mixed types into a single model. For growth curve models, the latent process reflects the 'true' growth.

Explore related subjects

Air PollutantsAsthmaBritish ColumbiaBronchitisData Interpretation, StatisticalEar DiseasesHumansLongitudinal StudiesModels, BiologicalMultivariate AnalysisPoliomyelitisRespiration DisordersSulfurJournal ArticleReview

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

B Jorgensen, M Tsao. Dispersion models and longitudinal data analysis.. https://doi.org/10.1002/(sici)1097-0258(19990915%2F30)18%3A17%2F18%3C2257%3A%3Aaid-sim253%3E3.0.co%3B2-m

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2026

Stable chloroform emissions in southeastern China: insights from recent observations.

Chloroform (CHCl3) is a short-lived ozone-depleting substance not currently regulated under the Montreal Protocol. Due to the unique meteorological conditions in East Asia, CHCl3 emitted in this region has a greater potential to reach the stratosphere and contribute to ozone depletion. As an essential component of national CHCl3 emissions, southeastern China has attracted increasing attention. However, long-term observational data in this region remain relatively scarce, with no updates since 2020. In this study, we continuously measured atmospheric CHCl3 concentrations at a remote monitoring station in southeastern China from 1 June 2023 to 31 May 2025. Frequent concentration enhancements were observed during the monitoring period, with mixing ratios ranging from 9.3 to 134.1 ppt and an average value of 36.8 ± 19.7 ppt. Back-trajectory analysis indicated that air masses associated with elevated CHCl3 levels primarily originated from coastal industrial provinces in eastern China. Using the Potential Source Contribution Function and Concentration Weighted Trajectory methods, we identified the Yangtze River Delta (Jiangsu, Anhui, and Zhejiang Provinces) and Jiangxi Province as dominant source regions. Emissions of CHCl3 in southeastern China were estimated using the Tracer Ratio Method to be approximately 30.1 ± 5.3 Gg/yr from 2023-06-01 to 2025-05-31, indicating overall stability relative to earlier estimates and no apparent upward trend. These findings provide updated insights into the current status of CHCl3 emissions in southeastern China and highlight the need for continued monitoring and emission assessment of CHCl3 in East Asia, given its unregulated status and implications for ozone layer recovery.

Air Pollutants↗
PubMed2000

Insoluble components of concentrated air particles mediate alveolar macrophage responses in vitro.

We sought to characterize the bioactive constituents of concentrated ambient air particles (CAPs) through correlation of alveolar macrophage (AM) biological responses (production of TNF, macrophage inflammatory protein (MIP-2), nitrite; cell viability) to components of particle samples. CAPs samples collected on different days showed a range of bioactivity and a strong correlation was found between AM cytokine release and increased AM light scatter, a flow cytometric measure of relative particle load. Evaluation of soluble and insoluble fractions of CAPs suspensions indicate that 1) most biological effects on AMs are mediated by insoluble components and certain particle adsorbed factors such as endotoxin; 2) the variable bioactivity of CAPs collected on different days arises primarily from differences in the relative proportion of insoluble and soluble mass present in particle suspensions; and 3) the activation state of the AM influences which insoluble components are most bioactive. Use of endotoxin neutralizing agents (e.g., polymyxin-B) showed particle-adsorbed endotoxin in CAPs suspensions causes activation of normal (control) AMs while other (nonendotoxin) components are predominantly responsible for the enhanced cytokine release observed by primed AMs incubated with CAPs. The AM biological response did not correlate with any of a panel of elements quantified within insoluble CAPs samples (Al, Cd, Cr, Cu, Fe, Mg, Mn, Ni, S, Ti, V). These data demonstrate an important role for cell activation and phagocytosis of insoluble particulate matter in the response of AMs to CAPs suspensions.

Air Pollutants↗
PubMed2000

Estimating particulate matter-mortality dose-response curves and threshold levels: an analysis of daily time-series for the 20 largest US cities.

Numerous studies have shown a positive association between daily mortality and particulate air pollution, even at concentrations below regulatory limits. These findings have motivated interest in the shape of the exposure-response relation. The authors have developed flexible modeling strategies for time-series data that include spline and threshold exposure-response models; they apply these models to daily time-series data for the 20 largest US cities for 1987-1994, using the concentration of particulate matter <10 microm in aerodynamic diameter (PM10) as the exposure measure. The spline model showed a linear relation without indication of threshold for PM10 and relative risk of death for all causes and cardiorespiratory causes; by contrast, for other causes, the risk did not increase until approximately 50 microg/m3 PM10. For all-cause mortality, a linear model without threshold was preferred to the threshold model and to the spline model, using the Akaike information criterion (AIC). The findings were similar for cardiovascular and respiratory deaths combined. By contrast, for causes other than cardiovascular and respiratory, a threshold model was more competitive with a threshold value estimated at 65 microg/m3. These findings indicate that linear models without a threshold are appropriate for assessing the effect of particulate air pollution on daily mortality even at current levels.

Air Pollutants↗