Latent profile analysis is a form of finite mixture modeling, a family of statistical methods designed to uncover hidden subgroups within a population. Instead of assuming that all smokers come from one homogeneous distribution, the method tests whether the observed pattern of responses across several indicators can be better explained by a small number of underlying groups, each with its own characteristic response pattern. In this study, the indicators were five in number: a measure of psychological dependence on smoking, the number of cigarettes smoked per day, motives for initiating smoking, motives for maintaining the habit, and the severity of withdrawal symptoms. The modeling, conducted in the R statistical environment using Gaussian finite mixture models, compared solutions with increasing numbers of profiles and evaluated them with standard fit criteria, including entropy-based measures of classification quality and likelihood-based information criteria, before settling on a four-profile solution as the best balance of parsimony and explanatory power.

The largest group by far, accounting for 58.9 percent of the sample, was labeled the Moderate Habitual profile. These smokers occupied the statistical middle ground across most indicators: moderate psychological dependence, moderate consumption, moderate motives, and moderate withdrawal. They represent the everyday image of the habitual smoker, someone whose habit is woven into daily routines without reaching the extremes seen elsewhere in the data. The second-largest group, at 18.5 percent, was the Low-Engagement profile. These smokers scored lowest on nearly every indicator, showing weak psychological dependence, low consumption, and mild withdrawal symptoms. Crucially, they also reported the highest intention to quit smoking of any group, a finding that suggests their lighter engagement with the habit is accompanied by a genuine openness to giving it up.

The two smaller profiles are where the study’s most striking insights lie. The Motivation-Driven profile, comprising 16.8 percent of smokers, displayed the highest levels of psychological dependence, the strongest smoking motives, and the most severe withdrawal symptoms of any group. What defines these smokers is not sheer volume of consumption but the intensity of the psychological and motivational machinery behind their habit. They smoke for reasons, and those reasons grip them tightly. By contrast, the Habitual Heavy profile, at just 5.8 percent of the sample, reported the highest number of cigarettes smoked per day, even though their psychological dependence and withdrawal symptoms did not top the scale. For this small group, smoking appears to be primarily a matter of deeply entrenched behavioral volume rather than acute psychological craving.

This dissociation between consumption and psychological dependence is perhaps the study’s most consequential finding. Public health screening and cessation triage have often leaned on cigarette counts as a proxy for how hard it will be for someone to quit. The four-profile solution shows that this proxy can mislead. A Motivation-Driven smoker who consumes a moderate number of cigarettes may, in psychological terms, be more firmly hooked than a Habitual Heavy smoker who burns through far more tobacco. Conversely, the Heavy smoker’s problem may be less about craving and more about the sheer density of smoking cues and routines packed into the day. If these distinctions hold up in future research, they imply that a one-size-fits-all cessation program, or one calibrated only to consumption levels, will fit poorly for at least two of the four profiles.

The study also examined which background characteristics predicted membership in each profile, using auxiliary variable techniques that relate covariates to latent classes without allowing them to distort the underlying classification. Five factors emerged as significantly associated with profile membership: family tobacco use, the presence of a chronic disease, sex, age, and cessation intention. Growing up or living with family members who smoke has long been recognized as a powerful social determinant of tobacco behavior, and here it helped distinguish between profiles, likely reflecting both early exposure to smoking norms and continued environmental reinforcement of the habit. Chronic disease status also differentiated the groups, a reminder that health conditions often intersect with smoking behavior in complex ways, sometimes motivating reduced smoking and sometimes complicating cessation efforts.

The roles of sex and age echo a broader literature on how smoking trajectories unfold across the life course. Smoking patterns established in adolescence and early adulthood tend to persist, and the demographic correlates identified in this study are consistent with the idea that profile membership is not randomly distributed but shaped by developmental and social context. Cessation intention, meanwhile, stood out as both an outcome and a predictor: Low-Engagement smokers, with their weak dependence, expressed the strongest desire to quit, while other profiles, presumably those more deeply entrenched, showed less intention. This gradient suggests that intention to quit is not simply a matter of willpower distributed unevenly across individuals, but is partly a function of the psychological structure of the dependence itself.

The methodological rigor of the study deserves attention. The author followed established guidance for latent profile analysis, including a priori consideration of statistical power for mixture models, careful model comparison across competing numbers of profiles, and the use of three-step approaches for relating auxiliary variables to profile membership, which help avoid the bias that arises when covariates are folded directly into the classification model. Multiple-testing concerns were addressed with false discovery rate control, and continuous variables such as age were handled without arbitrary dichotomization, a practice known to distort odds ratios. Supplementary materials are publicly archived on the Open Science Framework, and the study protocol received institutional ethics approval, with informed consent obtained from all participants. Raw participant-level data remain restricted for privacy reasons but are available on reasonable request.

The context of Türkiye gives the findings particular public health weight. National health survey data indicate that a substantial share of Turkish adults smoke, and tobacco-attributable disease imposes heavy costs on households and the health system. Globally, smoking remains one of the leading preventable causes of death, and cessation interventions, from nicotine replacement therapy to text-message and app-based support programs, have shown measurable but uneven success. The heterogeneity documented in this study offers one plausible explanation for that unevenness: interventions designed for the average smoker may resonate with the Moderate Habitual majority while missing the specific psychological levers that drive Motivation-Driven smokers or the behavioral density that defines Habitual Heavy smokers.

The author is careful to frame the work as a foundation rather than a prescription. The profiles were identified in a cross-sectional sample, so the analysis captures smokers at a moment in time rather than tracking how individuals move between profiles over months and years. Whether a Motivation-Driven smoker gradually becomes a Low-Engagement one, or whether profile membership is stable, is a question for longitudinal designs. Nor can cross-sectional data establish causal direction between profile characteristics and variables such as cessation intention. Still, the study makes a compelling case that the phenotype of smoking is multidimensional, and that the dimensions do not always move together. As cessation science moves toward more personalized approaches, the four profiles identified here, Low-Engagement, Moderate Habitual, Motivation-Driven, and Habitual Heavy, offer a concrete, data-driven vocabulary for thinking about smokers not as points on a single severity scale, but as members of distinct populations with distinct paths out of the habit.

Subject of Research: Latent profile analysis of multidimensional cigarette smoking profiles among adult smokers in Türkiye

Article Title: Multidimensional Smoking Profiles Defined by Psychological Dependence, Smoking Motives, Cigarette Consumption, and Withdrawal Symptoms: A Latent Profile Analysis

Article References: Bulut, A. (2026). Multidimensional Smoking Profiles Defined by Psychological Dependence, Smoking Motives, Cigarette Consumption, and Withdrawal Symptoms: A Latent Profile Analysis. International Journal of Mental Health and Addiction. https://doi.org/10.1007/s11469-026-01738-9

Image Credits: AI Generated

DOI: 10.1007/s11469-026-01738-9

Keywords: smoking, tobacco dependence, latent profile analysis, psychological dependence, smoking motives, withdrawal symptoms, smoking cessation, Türkiye, nicotine, public health, addiction research, Multidimensional