Late relapse/sustained abstinence among former smokers: a longitudinal study

Abstract

Background. Relapse remains the most refractory aspect of nicotine dependence, but little is known about relapse that occurs after a period of prolonged abstinence (i.e., “late relapse”).

Method. The current study examined the prevalence and predictors of late relapse/sustained abstinence in a community-based, longitudinal cohort of former smokers surveyed at baseline and again 4 years later (n = 1143).

Results. Late relapse rates declined precipitously and sustained abstinence increased as the duration of abstinence at baseline increased. Although the risk for late relapse appeared to be small when using a point prevalence definition of smoking, that risk increased substantially if fluctuations in smoking behavior over time were considered (i.e., many former smokers relapsed and then regained abstinence during the 4-year study period). Although there were a number of significant prospective predictors of late relapse/sustained abstinence, they did not retain their predictive ability after controlling for baseline duration of abstinence.

Conclusion. Although late relapse rates in the general population of former smokers appear low, and particularly so for individuals who have been abstinent for long periods of time, fluctuations between abstinence and smoking over time are not uncommon. How these fluctuations influence health or sustained abstinence over time is unclear. Because late relapse rates are generally low, interventions are likely to require careful targeting of at-risk individuals if they are to be efficacious and cost-effective.

Introduction

Relapse remains the most refractory aspect of nicotine dependence, and some researchers have suggested that there is no “safe” point at which former smokers are no longer at risk for relapse [1], [2]. Unfortunately, little is known about relapse among long-term former smokers [3], [4]. Because smoking is the leading cause of preventable morbidity and mortality in the United States [5], “late relapse” could have important implications for public health (“late relapse” is defined here as relapse after a period of sustained abstinence of at least 6–12 months). However, if the incidence of late relapse is low, and this is particularly likely to be true as the duration of abstinence increases, interventions targeted at the entire population of former smokers may not be necessary, feasible, or cost-effective. If interventions to prevent late relapse are in fact needed, it is likely that they would have to be carefully targeted at identifiable high-risk individuals or populations to warrant scarce public health resources. Thus, valid and reliable predictors of risk would be necessary to target high-risk groups. The purpose of the current study was to examine the prevalence and predictors of late relapse/sustained abstinence in a community-based, longitudinal sample of employed adults.

The few data available from community-based samples suggest that late relapse rates per year are typically 5% or less. In a probability sample of Michigan adults aged 18 and over, relapse rates per year averaged 4.8% among former smokers abstinent for 2 to 10 years and 1.2% among former smokers abstinent for more than 10 years [6]. In a longitudinal study of volunteers attempting unaided smoking cessation, relapse rates per year averaged 3% among individuals who had been abstinent for a year [7]. Among participants in the 1990 California Tobacco Survey, a random-digit dialed telephone survey of California adults, 10% of former smokers abstinent for 1–2 years relapsed within the subsequent 2 years—an average of 5% per year [8]. More recently, late relapse rates from the VA Normative Aging Study were 2–4% per year among elderly male veterans from the Boston area who had been abstinent for 2–6 years [9]. Among former smokers who volunteered for a relapse prevention study, late relapse rates over a 1-year period were approximately 5% among former smokers abstinent for 7–18 months at baseline and 3–4% among individuals abstinent for over 18 months [10].

It is noteworthy that, with few exceptions, only a small number of studies have investigated the occurrence of late relapse among former smokers who have been abstinent for 5 or more years. Thus, representative data on late relapse/sustained abstinence are needed across a spectrum of former smokers who differ in their duration of abstinence, including former smokers who have been abstinent for many years. Furthermore, previous studies have typically not addressed fluctuations in smoking behavior over time. In other words, because late relapse rates are generally based on point prevalence or continuous abstinence measures, there may be considerable relapse and recovery over time that is not accounted for in examining those figures.

A host of variables have been shown to predict relapse that occurs within the first 6–12 months after quitting including lower levels of education, motivation/readiness to change, self-efficacy/confidence, and social support; and higher levels of nicotine dependence, alcohol use, negative affect/stress, exposure to other smokers, and psychiatric comorbidity [7], [11], [12], [13], [14], [15]. Variables found to predict relapse among former smokers who have been abstinent for at least 6–12 months are generally similar and include younger age [8], [16]; gender (more relapse among women, [17]); use of other substances such as alcohol, caffeine, cigars, and pipes [9]; other people in the household smoking [18]; and nicotine dependence [19].

The most potent predictor of late relapse, however, is duration of abstinence [8], [10]. In the California Tobacco Survey, individuals abstinent for greater than 3 years were over 100 times more likely to remain abstinent than were former smokers who had been abstinent for less than 3 months, individuals abstinent for 1–3 years were 19 times more likely to remain abstinent, and individuals abstinent for 3–12 months were six times more likely to remain abstinent [8]. Not surprisingly, risk for late relapse decreases dramatically over time, and a challenge for researchers is to determine whether valid and reliable risk factors can be identified that independently predict late relapse over and above the effects of abstinence duration alone.

The current study examined the prevalence and predictors of late relapse/sustained abstinence in a community-based longitudinal sample of employed adults. Participants were former smokers who were surveyed at baseline and again 4 years later. Potential predictors of late relapse/sustained abstinence included demographics (age, gender, race/ethnicity, marital status, education), environmental variables (others in the household smoking, feeling bothered by smoke at work), smokeless tobacco use, alcohol use, self-rated health, and confidence in remaining abstinent. This is one of the few studies to prospectively examine late relapse/sustained abstinence in a large, community-based sample of former smokers who differ in their duration of abstinence.