Detecting longitudinal patterns of daily smoking following drastic cigarette reduction

Abstract

To enhance prolonged smoking cessation or reduction, a better understanding of the process of change is needed. This study examines daily smoking rates following the end of an intensive smoking reduction program originally designed to evaluate the relationship of tobacco biomarkers with reduced levels of smoking. A novel pattern-oriented approach called time-series-based typology is used to detect homogeneous smoking patterns in time-intensively (i.e., 40 occasions) observed smokers (n=57), who were predominantly Caucasian (94.7%), male (52.6%), and on average 47.9 years old (SD=11.3). The majority of the smokers exhibited a change in their daily smoking behavior over the course of 40 days with 47.4% increasing and 40.4% decreasing the number of cigarettes smoked per day, which is contrary to the results a group level approach would have found. Very few smokers (12.3%) maintained their average smoking rate, and exhibited an externally controlled smoking pattern. Trajectory type could be predicted by temporally proximal motivation and selfefficacy variables ((F(4, 106)=3.46, p=.011, η 2 =.115), underscoring their importance in maintaining reduced smoking rates. Time series-based typology demonstrated good sensitivity to the identification of meaningfully different trajectories. Keywordstime series-based typology; time series analysis; harm reduction; longitudinal smoking patterns; idiographic research A new wave of behavior change intervention research is beginning that seeks to usher a new generation of more effective, far-reaching, and sustainable interventions into existence (Orleans, 2005; S. D. Solomon, 2005). The creation and dissemination of innovative methodological tools that allow for the analysis of complex longitudinal data is an important part of this new initiative. In this paper, we present an innovative approach to longitudinal data that is based on traditionally used analyses and is thus very accessible to applied researchers. In particular, our paper focuses on detecting longitudinal patterns of daily smoking following drastic cigarette reduction. With most cigarette smokers relapsing within three months of a cessation attempt (Fiore, Smith, Jorenby, & Baker, 1994;Shiffman, 1993), there is a need to improve our understanding of the process of behavior change following smoking reduction Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. and cessation and underlying relapse. To this end, our paper presents the largest time series study of smoking behaviors to date. It is also our starting point for showcasing an innovative and easy to use analytical to...

Time Series Analysis of Treatment Adherence Patterns in Individuals with Obstructive Sleep Apnea

“…For most participants, the level of use on any given night is not affected by use levels on previous occasions. In contrast, data from daily smoking behavior typically reports autocorrelations of 0.50 and above [26,27].…”

Identifying Longitudinal Patterns for Individuals and Subgroups: An Example with Adherence to Treatment for Obstructive Sleep Apnea

“…In this article we are referring to this general approach as a Typology of Temporal Patterns (TTP). There are many examples of TTP in the literature including studies of patterns of daily smoking (Chandra, Shiffman, Scharf, Dang, & Shadel, 2007;Hoeppner, Goodwin, Velicer, Mooney, & Hatsukami, 2008), daily alcohol use (Harrington, Velicer, & Ramsey, 2014), and OSA (Aloia et al, 2008).…”

“…In traditional cluster analysis, homogenous subgroups are created from a number of variables measured on one occasion (Everitt, Landau, Leese, & Stahl, 2011). In dynamic cluster analysis, homogenous subgroups are created from one variable measured across a number of occasions (Hoeppner et al, 2008;Norman, Velicer, Fava, & Prochaska, 1998;Prochaska, Velicer, Guadagnoli, Rossi, & DiClemente, 1991). In the present study, homogenous subgroups were identified by clustering hours of PAP use over time.…”

Daily College Student Drinking Patterns Across the First Year of College

“…To identify subgroups of participants with similar longitudinal drinking patterns, we used time series-based typology (Hoeppner et al, 2008), a process that combines time series analysis (TSA) and cluster analysis. We chose this approach because our interest lies in patterns of drinking.…”