Abstract
Background and Aims
The evidence concerning the relationships between loneliness, gambling to escape, and problem gambling is mixed. This study aimed to investigate how loneliness relates to gambling to escape and gambling problems using a longitudinal approach.
Method
This population-based, longitudinal study included five time points, with data having been collected between April 2021 (T1) and April-May 2023 (T5). Participants were 18–75-year-old Finnish residents. Only those who had taken part in the study at each time point (T1–T5) and had gambled at least once a month at some point in the follow-up period were included in the study (n = 612; 54.58% male; Mage = 51.85 years). Loneliness was measured with the UCLA 3-Item Loneliness Scale, and gambling to escape was measured with three questions concerning negative escapism taken from the Motivations to Play Inventory. Problem gambling was measured using the Problem Gambling Severity Index (PGSI). Random intercept cross-lagged panel modeling was used to analyze the relationships.
Results
Gambling problems predicted future loneliness on a within-person level, but loneliness did not predict future gambling problems. Also, gambling to escape predicted future gambling problems on a within-person level. On a between-person level, loneliness, gambling to escape and gambling problems were positively correlated.
Discussion and Conclusion
Gambling problems may predispose individuals to future loneliness. However, the relatively small effects observed indicate that individual differences play a significant role in this regard.
Introduction
The global expansion of the gambling industry and the possibilities of online gambling have made gambling widely accessible and popular (Allami et al., 2021; Sulkunen et al., 2021). For most people, gambling is harmless recreation, but for some, gambling develops into a problem, with various associated social, financial, and psychological harms (Hilbrecht et al., 2020). Extensive research has found that psychosocial problems, such as loneliness, often co-occur with addictive behaviors (Savolainen, Oksanen, Kaakinen, Sirola, & Paek, 2020; van der Maas, 2016; Zilberman, Yadid, Efrati, & Rassovsky, 2020). Similarly, gambling problems have been found to be associated with loneliness (Sirola, Kaakinen, Savolainen, & Oksanen, 2019; Sirola, Nyrhinen, & Wilska, 2023; Vuorinen et al., 2021). However, the evidence concerning the association between loneliness and gambling problems is mixed (Nordmyr & Forsman, 2020). This study's purpose is to develop a nuanced understanding of the role of loneliness in gambling problems by investigating the relationships between loneliness, gambling to escape and problem gambling, using a longitudinal study design.
Loneliness is a subjective emotional state characterized by feelings of isolation and a lack of meaningful connections (Perlman & Peplau, 1982). It is often described as a perceived discrepancy between the desired and actual levels of social interaction (Cacioppo & Cacioppo, 2012). As the need to belong is a fundamental human motivation (Baumeister & Leary, 1995), it is no wonder that loneliness has been associated with several negative outcomes, such as depression and other mental health problems, chronic illnesses, and addictive behaviors, including gambling problems (Mushtaq, Shoib, Shah, & Mushtaq, 2014; Park et al., 2020; Theeke, 2010). Some studies indicate that loneliness predicts gambling problems (Castrén et al., 2013; Edgren, Castren, Jokela, & Salonen, 2016; McQuade & Gill, 2012), whereas other studies suggest that gambling problems can cause loneliness when people with gambling problems attempt to hide their problematic behavior and thus avoid social stigma or negative reactions from loved ones (Dabrowska & Wieczorek, 2020). As the direction of causality is unclear, more longitudinal research is needed.
Gambling to escape is one of the basic gambling motives that has been recognized in several studies (Binde, 2013; Dechant, 2014; Francis, Dowling, Jackson, Christensen, & Wardle, 2015; Wardle et al., 2011), and it refers to distracting oneself from daily stresses, negative emotions and thoughts. In the previous literature, gambling to escape has been consistently associated with gambling problems (Alaba-Ekpo, 2024; Neophytou, Theodorou, Artemi, Theodorou, & Panayiotou, 2023). The escape motive plays an important role in the Pathways Model developed by Blaszczynski and Nower (2002). According to the model, those with gambling problems are not a homogenous group: instead, they can be divided into three subgroups (Kurilla, 2021; Nower, Blaszczynski, & Anthony, 2022). One of these subtypes, emotionally vulnerable gamblers, is characterized by a history of mood and anxiety disorders, hurtful past experiences, substance abuse, and the use of gambling as a coping mechanism to deal with negative emotions, which ultimately leads to gambling problems. The hypothesis linking dysfunctional emotion regulation and coping strategies to gambling problems has received further support (Bonnaire et al., 2022; Marchica, Keough, Montreuil, & Derevensky, 2020; Rogier & Velotti, 2018). Given the model described above, it is possible that loneliness drives some individuals to use gambling to alleviate the hurtful experience of loneliness, which can evolve into gambling problems. For example, Sirola et al. (2019) found that loneliness moderated the effect between gambling problems and online gambling community participation. Moreover, Holdsworth, Hing, and Breen (2012) found that women were more likely than men to use gambling to avoid feelings of loneliness. Escaping from feelings of loneliness, social isolation, and loss seems to be an important motivator for gambling among elderly as well (Botterill, Gill, McLaren, & Gomez, 2016; Martin, Lichtenberg, & Templin, 2011). There is also some evidence that the stresses associated with the COVID-19 pandemic, such as social isolation and mental health issues, contributed to people's gambling problems (Masaeli & Farhadi, 2021). To our knowledge, only one prior study has investigated whether coping strategies moderate the relationship between loneliness and problem gambling among students (Hum & Carr, 2018), but that study failed to find support for the hypothesis. However, that study was not nationally representative, as it used a convenience sample of students, and the study design was cross-sectional. Thus, more longitudinal research is needed.
Gambling problems may also predispose individuals to loneliness and social isolation, as gambling problems often cause shame, guilt, and depression (Hing, Nuske, Gainsbury, & Russell, 2016; Pchajek, Edgerton, Sanscartier, & Keough, 2023). Hiding a gambling problem may further distance oneself from other people (Dabrowska & Wieczorek, 2020). Gambling is also increasingly shifting to online platforms, and these modern forms of gambling may contribute to loneliness given that online gambling is typically an individual activity (Sirola et al., 2023). Moreover, individuals with gambling problems commonly suffer from additional addictive behaviors and psychosocial comorbidities, such as depression and anxiety, which can exacerbate their social isolation (Macía, Jauregui, & Estevez, 2023; Walther and Alhughe, 2012).
Given that gambling problems are often associated with financial and relationship difficulties, as well as feelings of guilt and shame (de Ridder & Deighton, 2022; Marko, Thomas, Pitt, & Daube, 2023), individuals with problematic gambling may turn to gambling to alleviate the stress and difficult emotions arising from these situations. This, in turn, may create a vicious cycle in which gambling itself causes distress and acts as a distraction from negative feelings. Some gamblers with escapist motives even report seeking a dissociative state of mind, or “dark flow” (Kruger et al., 2020; Wood & Griffiths, 2007). Choosing this kind of activity as a way to cope with distress may isolate individuals and potentially contribute to loneliness.
The present study
The aim of this longitudinal study is to investigate the dynamic relationships between loneliness, the escape motive, and gambling problems. As loneliness is a painful experience, it may lead some people to engage in gambling to cope with the distressing feelings that arise due to social isolation (Edgren et al., 2016; Holdsworth et al., 2012; Sirola et al., 2019). With time, this can lead to gambling problems. However, it is also common that individuals with gambling problems attempt to conceal their situation and thus avoid social stigma and negative reactions (Dabrowska & Wieczorek, 2020), which may isolate them from other people and cause loneliness. Several studies have also found strong evidence that using gambling as a coping mechanism can offer temporary relief from distressing feelings and thoughts. However, in the long term, this leads to gambling problems (Alaba-Ekpo, 2024; Neophytou et al., 2023). Finally, escapist gambling behavior may isolate individuals from social connections especially if they use gambling to enter a dissociative state of mind (Kruger et al., 2020; Wood & Griffiths, 2007). In addition to these dynamic associations, it is possible that the connections between loneliness, gambling to escape, and gambling problems reflect more stable individual differences, such as differences in personality (see e.g. Blaszczynski & Nower, 2002).
In this study, we examine cross-lagged paths to clarify the causal relationships between loneliness and problem gambling, as well as how the escapist motivation to gamble contributes to this. To obtain more accurate results, we separate the time-varying within-person variance from the stable, trait-like between-person variance. Within-person variance represents changes in an individual's measurement scores over time relative to their own mean level and captures state-like dynamic changes that occur within an individual from one time point to another. This variance is crucial for understanding how deviations from an individual's average levels over time might predict subsequent changes in another variable. Between-person variance, on the other hand, captures the extent to which individuals differ from one another in their average levels of the variables measured. This variance is more stable over time and reflects trait-like characteristics (Berry & Willoughby, 2017).
Based on the previous literature, we present the following hypotheses on the within-person level:
Loneliness predicts future gambling to escape.
Loneliness predicts future gambling problems.
Gambling problems predict future loneliness.
Gambling problems predict future gambling to escape.
Gambling to escape predicts future gambling problems.
Gambling to escape predicts future loneliness.
On the between-person level, we hypothesize that loneliness, gambling to escape, and gambling problems are positively correlated (H7).
Method
Participants and procedure
A nationwide sample of Finnish residents aged 18–75 years (N = 1,530, 51.27% male, 48.43% female, 0.29% other gender) was recruited from Mainland Finland. The data were collected through a panel administered by the European data collection company Norstat. The data collection was conducted at five timepoints, which occurred every 6 months, beginning in April 2021 (T1, response rate 34.60%). The first follow-up survey took place in October–November 2021 (T2, n = 1,198, response rate 78.30%), followed by the second in April–May 2022 (T3, n = 1,100, response rate 91.40%). The third follow-up survey was collected in October–November 2022 (T4, n = 1,008, response rate 91.60%) and the fourth was collected in April–May 2023 (T5, n = 934, response rate 93.02%). At each timepoint, researchers performed data quality checks to eliminate overtly biased response patterns from the final dataset. Only participants who took part at all timepoints (n = 812) and had gambled using at least one game type at least once a month during the 2.5 years of data collection were included in the final sample (n = 612, 40.00% of the original T1 sample, 54.58% male, 45.42% female, 0.16% other gender). The use of five time points allows for the examination of dynamics in gambling problems and other variables both in the short term and over the longer term. Spanning 2.5 years, the data allows us to observe meaningful changes and patterns across time.
According to the non-response analysis, those who took part at all five time points and had gambled at least once per month were older than the original T1 participants (mean age at T1 46.67 years vs. 51.85 years at T5). No major effects on drop out were found based on gender, geographical area, education, income, marital status or occupational status (see Oksanen, Mantere, Vuorinen, & Savolainen, 2022), although there were slightly fewer at-risk gamblers in the final sample compared to the original T1 sample (score of 5 or higher on the Problem Gambling Severity Index, T1 = 9.48% vs. T5 = 8.99%). Compared to the general Finnish population, this sample does not contain any major biases, except for the abovementioned drop-out among younger participants (Grönroos, Salonen, Latvala, Kontto, & Hagfors, 2024; Oksanen, et al., 2022), though analytical weights were used to correct for this bias.
Measures
Gambling problems
The Problem Gambling Severity Index (PGSI) was used to measure gambling problems (Currie, Casey, & Hodgins, 2010; Ferris & Wynne, 2001). The PGSI is widely regarded as a reliable measure of gambling problems (Miller, Currie, Hodgins, & Casey, 2013; Orford, Wardle, Griffiths, Sproston, & Erens, 2010). It consists of nine items measuring various dimensions of gambling problems over the past six months (e.g., “Have you felt that you might have a problem with gambling?”). Respondents assessed each item on a four-point scale (0 = never, 1 = sometimes, 2 = most of the time, 3 = almost always). Scores range from 0 to 27, with higher scores indicating more severe gambling problems. McDonald's omega coefficients showed excellent internal consistency for the scale (T1: ω = 0.94, T2: ω = 0.94, T3: ω = 0.94, T4: ω = 0.94, T5: ω = 0.94).
Loneliness
Loneliness was measured with the UCLA Three-Item Loneliness Scale (Hughes, Waite, Hawkley, & Cacioppo, 2004). These questions asked respondents to rate how often they experienced different forms of loneliness or social isolation (“How often do you feel isolated from other people?”, “How often do you feel lonely?”, and “How often do you feel left out?”). Items were assessed using a three-point scale (0 = Hardly ever, 1 = Occasionally, 2 = Often). Potential values for the loneliness variable ranged from 0 to 6, with higher values indicating a greater experience of loneliness. McDonald's omega indicated good internal consistency for the scale (T1: ω = 0.84, T2: ω = 0.83, T3: ω = 0.85, T4: ω = 0.84, T5: ω = 0.84).
Gambling to escape
Gambling to escape was measured using three questions concerning negative escapism derived from the Motivations to Play Inventory (Hagström & Kaldo, 2014; Jouhki & Oksanen, 2022; Yee, 2006). The items focus on the impact of various escape motives on gambling in the past six months (“How often in the last 6 months have you gambled to avoid thinking about some of your real-life problems or worries?”, “How often in the last 6 months have you gambled to avoid real-life social encounters or situations?”, and “How often in the last six months have you continued to gamble to avoid having to deal with everyday problems and conflicts?”). Responses were given on a five-point scale (0 = Never, 1 = Rarely, 2 = Sometimes, 3 = Often, 4 = Always). Potential values on the escape motive scale ranged from 0 to 12, with higher values indicating a higher prevalence of escape motives to gamble. The McDonald's omegas for the scale were as follows: T1: ω = 0.86, T2: ω = 0.86, T3: ω = 0.87, T4: ω = 0.84, T5: ω = 0.87.
Statistical analysis
We used the random-intercept cross-lagged panel model (RI-CLPM) to analyze the relationships between loneliness, gambling to escape and gambling problems at five time points. The RI-CLPM is an extension to the traditional cross-lagged panel model, which has been criticized for its inability to differentiate between the within-person and between-person changes in longitudinal panel studies which may lead to spurious conclusions (Lucas, 2023). The RI-CLPM has been designed to address this problem by adding a random intercept into the model (Hamaker, Kuiper, & Grasman, 2015; Mulder & Hamaker, 2021). The RI-CLPM decomposes the observed variance into stable, time-invariant between-person differences, as well as time-varying within-person differences. The between-person component reflects variance due to differences that exist between persons (e.g., people differ in terms of their average level of gambling problems), while the within-person component reflects variance due to changes which vary within individuals over time (e.g., an individual's deviations from their personal average level of gambling problems).
The descriptive analyses were performed with Stata software 18.0 (StataCorp, 2023), and RI-CLPM was performed with the lavaan package for R version 0.6–17 (Rosseel, 2012). We used a maximum likelihood estimator (MLE) with robust standard errors. We also performed square root transformations for loneliness, escape motive, and PGSI score to adjust for multivariate nonnormality. We built two models: in the first model (Model 1), no parameter constraints were set, whereas in the second model (Model 2), all the path coefficients and covariances of the error terms were constrained to be equal over time. To estimate the model fit, we report several fit indices: the Chi-squared test statistic, comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean squared residual (SRMR). The following cutoff criteria for good fit were used as suggested by Hu and Bentler (1999): above 0.90/0.95 for CFI, below 0.08 for SRMR, and below 0.06 for RMSEA. A change of less than −0.01 in the comparative fit index (CFI) was taken as evidence that the constrained model did not have a significantly worse fit than the unconstrained model (Putnick and Bornstein, 2016). If the constrained model had a significantly worse fit than the unconstrained model, this would imply that the time constraints are untenable and some kind of developmental process was taking place during the study (Mulder & Hamaker, 2021).
Ethics
The study underwent a review by The Academic Ethics Committee of the Tampere region and was approved before the initial data collection. The participants were provided with information about the survey's purpose and use, and their consent to participate was obtained via the completion of the entire survey. The identity of individual participants could not be discerned from the data, as the data collection company Norstat supplied only anonymized data to the researchers.
Results
Descriptive analyses
The descriptive statistics are reported in Table 1. Of the participants, 45.42% were female or other gender and 54.58% were male. Mean age was 51.85 years (SD = 13.97 years). The mean score for loneliness was 1.59 at T1, peaked at T3 (1.69) and was slightly lower at T5 (1.54). The mean score for the escape motive increased slightly from T1 (0.81) to T5 (0.89). The mean PGSI score decreased from T1 (1.39) to T5 (1.25).
Descriptive statistics of the variables (n = 612)
| Range | T1, M (SD) | T2, M (SD) | T3, M (SD) | T4, M (SD) | T5, M (SD) | |
| Lonelinessa | 0–6 | 1.59 (1.66) | 1.69 (1.65) | 1.69 (1.65) | 1.66 (1.65) | 1.54 (1.63) |
| Escapeb | 0–12 | 0.81 (1.69) | 0.85 (1.78) | 0.86 (1.83) | 0.92 (1.81) | 0.89 (1.86) |
| PGSIc | 0–25 | 1.39 (3.35) | 1.42 (3.32) | 1.42 (3.39) | 1.34 (3.26) | 1.25 (3.00) |
| Age | 18–75 | 51.85 (13.97) | – | – | – | – |
| Gender | % | n | ||||
| Men | 54.58 | 334 | ||||
| Women or other genderd | 45.42 | 278 |
aThe UCLA 3-Item Loneliness Scale.
bThe Motivations to Play Inventory – Negative escapism.
cProblem Gambling Severity Index (PGSI).
dOther gender n = 1.
The pairwise correlations between the main variables are presented in Table 2. The strongest correlations were between the same variables measured in different time, points indicating that the participants' scores for loneliness, escape motive, and gambling problems at preceding timepoints remained relatively stable at the followings time point. The correlations between the escape motive and loneliness were relatively low (r = 0.27–0.34, p < 0.001) while the correlations between the escape motive and gambling problems ranged from low to moderate (r = 0.39–0.53, p < 0.001) at various time points. The correlations between loneliness and PGSI ranged from 0.18 to 0.27 (p < 0.001).
Pairwise correlations of the variables in five time points (n = 612)
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | |
| 1. Loneliness T1 | 1 | |||||||||||||
| 2. Loneliness T2 | 0.73 | 1 | ||||||||||||
| 3. Loneliness T3 | 0.73 | 0.75 | 1 | |||||||||||
| 4. Loneliness T4 | 0.73 | 0.75 | 0.76 | 1 | ||||||||||
| 5. Loneliness T5 | 0.73 | 0.73 | 0.75 | 0.80 | 1 | |||||||||
| 6. Escape T1 | 0.32 | 0.27 | 0.28 | 0.30 | 0.29 | 1 | ||||||||
| 7. Escape T2 | 0.30 | 0.32 | 0.31 | 0.31 | 0.32 | 0.68 | 1 | |||||||
| 8. Escape T3 | 0.32 | 0.29 | 0.31 | 0.28 | 0.29 | 0.65 | 0.66 | 1 | ||||||
| 9. Escape T4 | 0.29 | 0.34 | 0.33 | 0.34 | 0.32 | 0.63 | 0.64 | 0.67 | 1 | |||||
| 10. Escape T5 | 0.30 | 0.29 | 0.26 | 0.31 | 0.30 | 0.63 | 0.63 | 0.59 | 0.70 | 1 | ||||
| 11. PGSI T1 | 0.27 | 0.26 | 0.24 | 0.23 | 0.25 | 0.48 | 0.47 | 0.43 | 0.46 | 0.41 | 1 | |||
| 12. PGSI T2 | 0.24 | 0.25 | 0.24 | 0.22 | 0.25 | 0.43 | 0.48 | 0.41 | 0.45 | 0.39 | 0.84 | 1 | ||
| 13. PGSI T3 | 0.21 | 0.22 | 0.21 | 0.21 | 0.21 | 0.39 | 0.40 | 0.47 | 0.43 | 0.36 | 0.76 | 0.76 | 1 | |
| 14. PGSI T4 | 0.23 | 0.23 | 0.21 | 0.22 | 0.24 | 0.41 | 0.44 | 0.46 | 0.53 | 0.46 | 0.78 | 0.81 | 0.78 | 1 |
| 15. PGSI T5 | 0.24 | 0.22 | 0.18 | 0.20 | 0.22 | 0.40 | 0.39 | 0.40 | 0.47 | 0.48 | 0.72 | 0.76 | 0.73 | 0.76 |
Note. Loneliness = The UCLA 3-Item Loneliness Scale. Escape = The Motivations to Play Inventory – Negative escapism. PGSI = Problem Gambling Severity Index (PGSI). All the correlations were statistically significant at p < 0.001.
The random-intercept cross-lagged panel models
Model 1 did not impose any constraints on the parameters, whereas in Model 2, all the autoregressive and cross-lagged paths and the covariances of the error terms were constrained to be equal over time. According to the fit statistics (Table 3), both Model 1 (χ2(48) = 63.675, p = 0.064; CFI = 0.998; RMSEA = 0.023; SRMR = 0.025) and Model 2 (χ2(84) = 145.195, p < 0.001; CFI = 0.992; RMSEA = 0.035; SRMR = 0.020) had excellent fits for the data, except for the Chi-squared test, which was statistically significant for Model 2. However, the Chi-squared test is known for being overly sensitive with large sample sizes and almost always rejects the model (Shi, Lee, & Maydeu-Olivares, 2019). The decrease in CFI was less than −0.01, indicating that the constrained model (Model 2) did not have a significantly worse fit than the unconstrained model (Model 1). Therefore, we present the results derived from Model 2 in Fig. 1.
The fit indices for model comparison
| Model fit indices | Model 1 | Model 2 |
| Chi-square | χ2(48) = 63.68, p = 0.064 | χ2(84) = 145.20, p < 0.001 |
| CFI | 0.998 | 0.992 |
| RMSEA | 0.023 | 0.035 |
| SMSR | 0.025 | 0.020 |
| CFI difference | −0.006 | |
Note. CFI = comparative fit index, RMSEA = Root Mean-Square Error of Approximation, SMSR = Standardized Root Mean Square Residual.
Random-intercept cross-lagged panel model (RI-CLPM) for loneliness (Loneliness), gambling to escape (Escape), and gambling problems (PGSI). The RI-CLPM decomposes the observed variance (rectangles) into latent between-person and within-person variables (ovals). A single-headed arrow indicates a causal relationship between two variables and a double-headed curved arrow represents a covariation between two variables. The numbers next to the arrows are path coefficients, which represent the strength and direction of the relationship. All the autoregressive and cross-lagged path coefficients were constrained to be equal over time.
Note. *p < 0.05, **p < 0.01, ***p < 0.001. Only three time points depicted for simplicity. The within-person correlations among latent variables in T1 and the covariance of the error terms are modeled in the analysis but not depicted in the figure for clarity.
Citation: Journal of Behavioral Addictions 2025; 10.1556/2006.2025.00025
According to the results, gambling problems predicted future loneliness (β = 0.069, p = 0.007) and gambling to escape predicted future gambling problems (β = 0.077, p = 0.047) at the within-person level, providing support for H3 and H5. However, the effects were relatively small, indicating that most of the variance was at the between-person level. Loneliness did not predict future gambling to escape or gambling problems (β = 0.001, p = 0.971). However, the autoregressive paths (i.e., variable predicting itself at the subsequent time points) for loneliness (β = 0.083, p = 0.024) and gambling to escape (β = 0.124, p = 0.005) were statistically significant. This means that both loneliness and gambling to escape are relatively persistent over time and their previous levels are good predictors of their future levels. The autoregressive path for gambling problems was not statistically significant, indicating that the variance was on the between-person level. On the between-person level, loneliness, gambling to escape, and gambling problems correlated positively with one another, lending support to H7. This means that on average, those with gambling problems tend to be lonely and gamble to escape. In addition, lonely individuals are more likely to gamble to escape.
We performed an additional sensitivity analysis using the full information maximum likelihood estimation and the same constraints. The results remained the same, except for autoregressive path for gambling problems, which was statistically significant (β = 0.12, p = 0.028), and the cross-lagged path from escape to loneliness, which was statistically significant but negative (β = −0.05, p = 0.018).
Discussion
This study investigated the dynamic relationships between loneliness, escapist gambling, and gambling problems using a longitudinal approach. Its analyses were based on a representative sample of Finnish adults. According to the results, gambling problems predict future loneliness on a within-person level. While the effect sizes were modest, these results suggest that gambling problems lead to loneliness, rather than vice versa. Additionally, gambling to escape predicted future gambling problems on a within-person level. On a between-person level, loneliness, gambling to escape, and gambling problems were positively correlated with one another, which indicates that on average, those who have gambling problems tend to be lonely and gamble to escape.
These results extend the large body of research indicating that gambling to escape plays a crucial role in gambling problems (Alaba-Ekpo, 2024; Neophytou et al., 2023). Our study provides support for the Pathways Model of problem gambling, which proposes that relying on gambling as a coping mechanism plays an essential role in the development of gambling problems (Blaszczynski & Nower, 2002; Nower et al., 2022). However, we could not confirm the hypothesis that loneliness leads directly to gambling to escape, which ultimately leads to gambling problems. One potential reason for this could be that our sample was demographically balanced and the levels of gambling to escape and gambling problems were relatively low, which is typical in addiction research. For example, a recent meta-analysis estimated that 8.7% of adults engaged in risky gambling, while only 1.41% of adults engaged in problematic gambling (Tran et al., 2024). The results might have been different if we had investigated a clinical sample. Individuals with strong problem gambling tendencies may be more likely to turn to escapist gambling when feeling lonely. However, it seems that loneliness, although it is a painful experience, does not universally drive gambling behavior.
On the other hand, our results provide support for the idea that gambling problems predict future loneliness. One potential explanation for this is that individuals with gambling problems often face shame and stigma, which may lead to secrecy, withdrawal from social activities, and distancing themselves from friends and family (Dabrowska & Wieczorek, 2020). Stigma and shame can prevent seeking help, and a delay in treatment may lead to the exacerbation of gambling problems (Suurvali, Cordingley, Hodgins, & Cunningham, 2009; Tavares, Martins, Zilberman, & el-Guebaly, 2002). Health-care professionals play a vital role in identifying and addressing gambling problems. By asking patients about gambling directly, they can initiate a conversation, provide support, and guide individuals toward the right services (Blank, Baxter, Woods, & Goyder, 2021). It is also important to take stigma and shame into account when developing treatments. For example, web-based counseling is becoming a prominent way to address stigma because of its ability to ensure confidentiality and anonymity (Rodda et al., 2013).
These results underscore the need for effective prevention and intervention strategies that promote healthy coping mechanisms and reduce gambling-related issues. Escape motives, gambling problems, and loneliness appear to be intertwined. For example, fostering productive, action-based coping skills could prevent future gambling problems. It is also important to note that loneliness, gambling to escape, and gambling problems were interrelated on a between-person level. This suggests that the connections between these phenomena are also explained by more stable traits, such as differences in personality (Blaszczynski and Nower, 2002). From the perspective of gambling problem interventions, this means that while the same individuals typically experience loneliness, gambling to escape, and gambling problems, effective treatment will need to be comprehensive and address not only these, but other factors as well.
As a starting point, efforts to address these issues should include evidence-based approaches, such as clinical interventions, therapy, and increased funding for mental health services that focus on gambling as a coping mechanism. As gambling increasingly moves to online environments, it is important to focus on technological solutions that can leverage these insights and the capability of the available technology. Such tools could utilize mobile apps offering real-time support or self-monitoring of gambling behaviors. These interventions could include features designed to help individuals recognize when their gambling behavior is driven by escapist motives, enabling them to observe and identify contextual factors more effectively. Mobile apps should offer users alternative coping strategies or immediate connection to support networks. Access to virtual peer support groups or opportunities for offline social interactions with such groups could be potential features of the app targeted to foster a sense of community and belonging. Collaboration between clinicians and technology developers is crucial in an effort to create tools that are evidence-informed and efficient in addressing both gambling and social challenges of those vulnerable to gambling problems. Overall, various treatment-focused applications and digital tools will benefit from the growing knowledge base on problem gambling.
Our study has multiple strengths, but certain limitations must be noted. First, the study was limited to Finland, representing a unique context in terms of gambling and social culture, and it is important to replicate the results in various cultural contexts. In addition, we relied on self-reported information which may introduce biases, such as recall bias or social desirability bias, especially when investigating topics that can be considered sensitive. While the current study involves a large sample and follows the same respondents for 2.5 years, it is based on the general population and not those with gambling problems per se, a study of whom might have yielded different types of relationships. These associations should also be investigated using clinical samples to provide further support and enable comparisons.
Additionally, our study did not consider different gambling types, underscoring the need to examine these phenomena across diverse gambling contexts. While evidence suggests that different gambling motives drive various forms of gambling types (e.g. Sundqvist, Jonsson, & Wennberg, 2016), those with problem gambling typically engage in a broad range of gambling activities rather than specializing in particular types (Downling et al., 2017; Leslie & McGrath, 2024). Future studies should also investigate the role of gender differences, as previous literature has indicated differences in gambling behavior and escapist motives between men and women (e.g., Jouhki, Savolainen, Hagfors, Vuorinen, & Oksanen 2024; Wenzel & Dahl, 2009). However, our robust methods and high-quality data suggest that among the general population, gambling problems are likely to lead to increased loneliness and that gambling motivated by the need to escape predicts future gambling problems.
Conclusion
This study showed that gambling problems predict future loneliness and gambling to escape predicts future gambling problems. The results were based on robust 5-time point data, and the study provided a much-needed longitudinal perspective on the topic. We believe that connections between loneliness, escapism, and gambling problems should be acknowledged by health professionals, counselors and legislators. Our results suggest that individuals who are lonely and gamble to escape are more likely to have gambling problems compared to others. Further research is needed to identify effective intervention methods. In addition, future studies on the topic should continue building longitudinal designs.
Funding sources
The study was funded by the Finnish Foundation for Alcohol Studies (Gambling in the Digital Age Project, 2021–2024, PI: A. Oksanen). In addition, HH has received a personal grant from the Finnish Foundation for Alcohol Studies.
Authors' contribution
HH: conceptualization, formal analysis, methodology, visualization, writing – original draft. MK: conceptualization, formal analysis, methodology, writing – original draft, visualization, writing – review & editing. IS: conceptualization, data curation, funding acquisition, investigation, writing – original draft, writing – review & editing. JV: conceptualization, writing – original draft, writing – review & editing. AO: conceptualization, writing – original draft, writing – review & editing, data curation, funding acquisition, investigation, supervision. All the authors had full access to the data used in this study and take full responsibility for the integrity of the data and the accuracy of the data analysis. All the authors have read and approved the final manuscript.
Conflict of interest
The authors declare no conflict of interest.
References
Alaba-Ekpo, O., Caudwell, K. M., & Flack, M. (2024). Examining the strength of the association between problem gambling and gambling to escape. A systematic review and meta-analysis. International Journal of Mental Health and Addiction. https://doi.org/10.1007/s11469-024-01354-5.
Allami, Y., Hodgins, D. C., Young, M., Brunelle, N., Currie, S., Dufour, M., … Nadeau, L. (2021). A meta‐analysis of problem gambling risk factors in the general adult population. Addiction, 116(11), 2968–2977. https://doi.org/10.1111/add.15449.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. https://doi.org/10.1037/0033-2909.117.3.497.
Berry, D., & Willoughby, M. T. (2017). On the practical interpretability of cross‐lagged panel models: Rethinking a developmental workhorse. Child Development, 88(4), 1186–1206. https://doi.org/10.1111/cdev.12660.
Binde, P. (2013). Why people gamble: A model with five motivational dimensions. International Gambling Studies, 13(1), 81–97. https://doi.org/10.1080/14459795.2012.71215.
Blank, L., Baxter, S., Woods, H. B., & Goyder, E. (2021). Should screening for risk of gambling-related harm be undertaken in health, care and support settings? A systematic review of the international evidence. Addiction Science & Clinical Practice, 16(1), 1–35. https://doi.org/10.1186/s13722-021-00243-9.
Blaszczynski, A., & Nower, L. (2002). A pathways model of problem and pathological gambling. Addiction, 97(5), 487–499. https://doi.org/10.1046/j.1360-0443.2002.00015.x.
Bonnaire, C., Devos, G., Barrault, S., Grall-Bronnec, M., Luminet, O., & Billieux, J. (2022). An empirical investigation of the Pathways Model of problem gambling through the conjoint use of self-reports and behavioural tasks. Journal of Behavioral Addictions, 11(3), 858–873. https://doi.org/10.1556/2006.2022.00055.
Botterill, E., Gill, P. R., McLaren, S., & Gomez, R. (2016). Marital status and problem gambling among Australian older adults: The mediating role of loneliness. Journal of Gambling Studies, 32(3), 1027–1038. https://doi.org/10.1007/s10899-015-9575-5.
Cacioppo, J. T., & Cacioppo, S. (2012). The phenotype of loneliness. European Journal of Developmental Psychology, 9(4), 446–452. https://doi.org/10.1080/17405629.2012.690510.
Castrén, S., Basnet, S., Salonen, A. H., Pankakoski, M., Ronkainen, J. E., Alho, H., & Lahti, T. (2013). Factors associated with disordered gambling in Finland. Substance Abuse Treatment, Prevention, and Policy, 8(1), 24. https://doi.org/10.1186/1747-597X-8-24.
Currie, S. R., Casey, D. M., & Hodgins, D. C. (2010). Improving the psychometric Properties of the problem gambling severity index. Canadian Consortium for Gambling Research.
Dabrowska, K., & Wieczorek, L. (2020). Perceived social stigmatisation of gambling disorders and coping with stigma. Nordisk Alkohol- & Narkotikatidskrift: NAT, 37(3), 279–297. https://doi.org/10.1177/1455072520902342.
de Ridder, B., & Deighton, R. M. (2022). The effect of shame and self-efficacy on gambling behaviour. Journal of Gambling Studies, 38(3), 1059–1073. https://doi.org/10.1007/s10899-021-10059-6.
Dechant, K. (2014). Show me the money: Incorporating financial motives into the gambling motives questionnaire. Journal of Gambling Studies, 30(4), 949–965. https://doi.org/10.1007/s10899-013-9386-5.
Dowling, N. A., Merkouris, S. S., Greenwood, C. J., Oldenhof, E., Toumbourou, J. W., & Youssef, G. J. (2017). Early risk and protective factors for problem gambling: A systematic review and meta-analysis of longitudinal studies. Clinical Psychology Review, 51, 109–124. https://doi.org/10.1016/j.cpr.2016.10.008.
Edgren, R., Castren, S., Jokela, M., & Salonen, A. H. (2016). At-risk and problem gambling among Finnish youth: The examination of risky alcohol consumption, tobacco smoking, mental health and loneliness as gender-specific correlates. NAD Nordic Stud Alcohol Drugs., 33(1), 61–79. https://doi.org/10.1515/nsad-2016-0005.
Ferris, J., & Wynne, H. (2001). The Canadian problem gambling index: Final report. Ottawa, Canada: Canadian Centre on Substance Abuse.
Francis, K. L., Dowling, N. A., Jackson, A. C., Christensen, D. R., & Wardle, H. (2015). Gambling motives: Application of the reasons for gambling questionnaire in an Australian population survey. Journal of Gambling Studies, 31, 807–823. http://dx.doi.org/10.1007/s10899-014-9458-1.
Grönroos, T., Salonen, A., Latvala, T., Kontto, J., & Hagfors, H. (2024). Suomalaisten rahapelaaminen 2023: Rahapelaaminen vähentynyt, peliongelma yleistynyt ja suhtautuminen pelaamiseen muuttunut. (Report No. 15/2024). Finnish Institute of Health and Welfare. https://urn.fi/URN:NBN:fi-fe2024040414663.
Hagström, D., & Kaldo, V. (2014). Escapism among players of MMORPGs-conceptual clarification, its relation to mental health factors, and development of a new measure. Cyberpsychology, Behavior and Social Networking, 17(1), 19–25. https://doi.org/10.1089/cyber.2012.0222.
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102–116. https://doi.org/10.1037/a0038889.
Hilbrecht, M., Baxter, D., Abbott, M., Binde, P., Clark, L., Hodgins, D. C., & Manitowabi, D. (2020). The conceptual framework of harmful gambling: A revised framework for understanding gambling harm. Journal of Behavioral Addictions, 9(2), 190–205. https://doi.org/10.1556/2006.2020.00024.
Hing, N., Nuske, E., Gainsbury, S. M., & Russell, A. M. T. (2016). Perceived stigma and self-stigma of problem gambling: Perspectives of people with gambling problems. International Gambling Studies, 16(1), 31–48. https://doi.org/10.1080/14459795.2015.1092566.
Holdsworth, L., Hing, N., & Breen, H. (2012). Exploring women's problem gambling: A review of the literature. International Gambling Studies, 12(2), 199–213.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. https://doi.org/10.1080/10705519909540118.
Hughes, M. E., Waite, L. J., Hawkley, L. C., & Cacioppo, J. T. (2004). A short scale for measuring loneliness in large surveys: Results from two population-based studies. Research on Aging, 26(6), 655–672. https://doi.org/10.1177/0164027504268574.
Hum, S., & Carr, S. M. (2018). Testing the emotional vulnerability pathway to problem gambling in culturally diverse university students. Journal of Gambling Studies, 34(3), 915–927. https://doi.org/10.1007/s10899-018-9753-3.
Jouhki, H., & Oksanen, A. (2022). To get high or to get out? Examining the link between addictive behaviors and escapism. Substance Use & Misuse, 57(2), 202–211. https://doi.org/10.1080/10826084.2021.2002897.
Jouhki, H., Savolainen, I., Hagfors, H., Vuorinen, I., & Oksanen, A. (2024). What are escapists made of, and what does it have to do with excessive gambling and gaming? International Journal of Mental Health and Addiction. https://doi.org/10.1007/s11469-024-01394-x.
Kruger, T. B., Dixon, M. J., Graydon, C., Stange, M., Larche, C. J., Smith, S. D., & Smilek, D. (2020). Using deliberate mind-wandering to escape negative mood states: Implications for gambling to escape. Journal of Behavioral Addictions, 9(3), 723–733. https://doi.org/10.1556/jba-9-723.
Kurilla, A. (2021). Is subtyping of gamblers based on the pathways model of problem and disordered gambling valid? A systematic review. Journal of Gambling Studies, 37(3), 983–1006. https://doi.org/10.1007/s10899-020-09995-6.
Leslie, R. D., & McGrath, D. S. (2024). A comparative profile of online, offline, and mixed-mode problematic gamblers’ gambling involvement, motives, and HEXACO personality traits. Journal of Gambling Studies, 40(1), 237–253. https://doi.org/10.1007/s10899-023-10193-3.
Lucas, R. E. (2023). Why the cross-lagged panel model is almost never the right choice. Advances in Methods and Practices in Psychological Science, 6(1). https://doi.org/10.1177/25152459231158378.
Macía, L., Jauregui, P., & Estevez, A. (2023). Emotional dependence as a predictor of emotional symptoms and substance abuse in individuals with gambling disorder: Differential analysis by sex. Public Health (London), 223, 24–32. https://doi.org/10.1016/j.puhe.2023.07.023.
Marchica, L. A., Keough, M. T., Montreuil, T. C., & Derevensky, J. L. (2020). Emotion regulation interacts with gambling motives to predict problem gambling among emerging adults. Addictive Behaviors, 106, 106378. https://doi.org/10.1016/j.addbeh.2020.106378.
Marko, S., Thomas, S. L., Pitt, H., & Daube, M. (2023). The lived experience of financial harm from gambling in Australia. Health Promotion International, 38(3). https://doi.org/10.1093/heapro/daad062.
Martin, F., Lichtenberg, P. A., & Templin, T. N. (2011). A longitudinal study: Casino gambling attitudes, motivations, and gambling patterns among urban elders. Journal of Gambling Studies, 27(2), 287–297. https://doi.org/10.1007/s10899-010-9202-4.
Masaeli, N., & Farhadi, H. (2021). Prevalence of internet-based addictive behaviors during COVID-19 pandemic: A systematic review. Journal of Addictive Diseases, 39(4), 468–488. https://doi.org/10.1080/10550887.2021.1895962.
McQuade, A., & Gill, P. (2012). The role of loneliness and self-control in predicting problem gambling behaviour. Gambling Research Journal of the National Association for Gambling Studies Australia, 24(1), 18–30.
Miller, N. V., Currie, S. R., Hodgins, D. C., & Casey, D. (2013). Validation of the problem gambling severity index using confirmatory factor analysis and rasch modelling: Validation of the problem gambling severity index. International Journal of Methods in Psychiatric Research, 22(3), 245–255. https://doi.org/10.1002/mpr.1392.
Mulder, J. D., & Hamaker, E. L. (2021). Three extensions of the random intercept cross-lagged panel model. Structural Equation Modeling, 28(4), 638–648. https://doi.org/10.1080/10705511.2020.1784738.
Mushtaq, R., Shoib, S., Shah, T., & Mushtaq, S. (2014). Relationship between loneliness, psychiatric disorders and physical health? A review on the psychological aspects of loneliness. Journal of Clinical and Diagnostic Research, 8(9), WE01–WE04. https://doi.org/10.7860/JCDR/2014/10077.4828.
Neophytou, K., Theodorou, M., Artemi, T.-F., Theodorou, C., & Panayiotou, G. (2023). Gambling to escape: A systematic review of the relationship between avoidant emotion regulation/coping strategies and gambling severity. Journal of Contextual Behavioral Science, 27, 126–142. https://doi.org/10.1016/j.jcbs.2023.01.004.
Nordmyr, J., & Forsman, A. (2020). A systematic review of psychosocial risks for gambling and problem gambling in the Nordic countries. Health, Risk & Society, 22(3–4), 266–290. https://doi.org/10.1080/13698575.2020.1796929.
Nower, L., Blaszczynski, A., & Anthony, W. L. (2022). Clarifying gambling subtypes: The revised pathways model of problem gambling. Addiction (Abingdon, England), 117(7), 2000–2008. https://doi.org/10.1111/add.15745.
Oksanen, A., Mantere, E., Vuorinen, I., & Savolainen, I. (2022). Gambling and online trading: Emerging risks of real-time stock and cryptocurrency trading platforms. Public Health (London), 205, 72–78. https://doi.org/10.1016/j.puhe.2022.01.027.
Orford, J., Wardle, H., Griffiths, M., Sproston, K., & Erens, B. (2010). PGSI and DSM-IV in the 2007 British gambling prevalence survey: Reliability, item response, factor structure and inter-scale agreement. International Gambling Studies, 10(1), 31–44. https://doi.org/10.1080/14459790903567132.
Park, C., Majeed, A., Gill, H., Tamura, J., Ho, R. C., Mansur, R. B., … McIntyre, R. S. (2020). The effect of loneliness on distinct health outcomes: A comprehensive review and meta-analysis. Psychiatry Research, 294, 113514. https://doi.org/10.1016/j.psychres.2020.113514.
Pchajek, J., Edgerton, J. D., Sanscartier, M., & Keough, M. (2023). Exploring the impact of shame and guilt on coping with gambling problems among emerging adult gamblers. Canadian Journal of Behavioural Science, 55(3), 177–188. https://doi.org/10.1037/cbs0000343.
Perlman, D., & Peplau, L. A. (1982). Loneliness: A sourcebook of current theory, research and therapy. Wiley.
Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Developmental Review, 41, 71–90. https://doi.org/10.1016/j.dr.2016.06.004.
Rodda, S. N., Lubman, D. I., Dowling, N. A., & McCann, T. V. (2013). Reasons for using web-based counselling among family and friends impacted by problem gambling. Asian Journal of Gambling Issues and Public Health, 3(1), 1–11. https://doi.org/10.1186/2195-3007-3-12.
Rogier, G., & Velotti, P. (2018). Conceptualizing gambling disorder with the process model of emotion regulation. Journal of Behavioral Addictions, 7(2), 239–251. https://doi.org/10.1556/2006.7.2018.52.
Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02.
Savolainen, I., Oksanen, A., Kaakinen, M., Sirola, A., & Paek, H. J. (2020). The role of perceived loneliness in youth addictive behaviors: Cross-national survey study. JMIR Mental Health, 22(1), e14035-. https://doi.org/10.2196/14035.
Shi, D., Lee, T., & Maydeu-Olivares, A. (2019). Understanding the model size effect on SEM fit indices. Educational and Psychological Measurement, 79(2), 310–334. https://doi.org/10.1177/0013164418783530.
Sirola, A., Kaakinen, M., Savolainen, I., & Oksanen, A. (2019). Loneliness and online gambling-community participation of young social media users. Computers in Human Behavior, 95, 136–145. https://doi.org/10.1016/j.chb.2019.01.023.
Sirola, A., Nyrhinen, J., & Wilska, T.-A. (2023). Psychosocial perspective on problem gambling: The role of social relationships, resilience, and COVID-19 worry. Journal of Gambling Studies, 39(3), 1467–1485. https://doi.org/10.1007/s10899-022-10185-9.
StataCorp (2023). Stata statistical software: Release 18. College Station, TX: StataCorp LLC.
Sulkunen, P., Babor, T. F., Cisneros Örnberg, J., Egerer, M., Hellman, M., Livingstone, C., … Rossow, I. (2021). Setting limits: Gambling, science and public policy—summary of results. Addiction, 116(1), 32–40. https://doi.org/10.1111/add.15241.
Sundqvist, K., Jonsson, J., & Wennberg, P. (2016). Gambling motives in a representative Swedish sample of risk gamblers. Journal of Gambling Studies, 32(4), 1231–1241. https://doi.org/10.1007/s10899-016-9607-9.
Suurvali, H., Cordingley, J., Hodgins, D. C., & Cunningham, J. (2009). Barriers to seeking help for gambling problems: A review of the empirical literature. Journal of Gambling Studies, 25(3), 407–424. https://doi.org/10.1007/s10899-009-9129-9.
Tavares, H., Martins, S. S., Zilberman, M. L., & el-Guebaly, N. (2002). Gamblers seeking treatment: Why haven’t they come earlier? Addictive Disorders & Their Treatment, 1(2), 65–69. https://doi.org/10.1097/00132576-200206000-00005.
Theeke, L. A. (2010). Sociodemographic and health-related risks for loneliness and outcome differences by loneliness status in a sample of U.S. older adults. Research in Gerontological Nursing, 3(2), 113–125. https://doi.org/10.3928/19404921-20091103-99.
Tran, L. T., Wardle, H., Colledge-Frisby, S., Taylor, S., Lynch, M., Rehm, J., … Degenhardt, L. (2024). The prevalence of gambling and problematic gambling: A systematic review and meta-analysis. The Lancet. Public Health, 9(8), e594–e613. https://doi.org/10.1016/S2468-2667(24)00126-9.
van der Maas, M. (2016). Problem gambling, anxiety and poverty: An examination of the relationship between poor mental health and gambling problems across socio-economic status. International Gambling Studies, 16(2), 281–295. https://doi.org/10.1080/14459795.2016.1172651.
Vuorinen, I., Oksanen, A., Savolainen, I., Sirola, A., Kaakinen, M., Paek, H. J., & Zych, I. (2021). The mediating role of psychological distress in excessive gambling among young people: A four-country study. International Journal of Environmental Research and Public Health, 18(13), 6973. https://doi.org/10.3390/ijerph18136973.
Walther, B., Morgenstern, M., & Hanewinkel, R. (2012). Co-occurrence of addictive behaviours: Personality factors related to substance use, gambling and computer gaming. European Addiction Research, 18(4), 167–174. https://doi.org/10.1159/000335662.
Wardle, H., Moody, A., Spence, S., Orford, J., Volberg, R., Jotangia, D., … Dobbie, F. (2011). British gambling prevalence survey 2010 prepared for: The gambling commission. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/243515/9780108509636.pdf.
Wenzel, H., & Dahl, A. (2009). Female pathological gamblers - A critical review of the clinical findings. International Journal of Mental Health and Addiction, 7(1), 190–202. https://doi.org/10.1007/s11469-008-9174-0.
Wood, R. T. A., & Griffiths, M. D. (2007). A qualitative investigation of problem gambling as an escape-based coping strategy. Psychology and Psychotherapy, 80(1), 107–125. https://doi.org/10.1348/147608306X107881.
Yee, N. (2006). Motivations for play in online games. Cyberpsychology & Behavior, 9(6), 772–775. https://doi.org/10.1089/cpb.2006.9.772.
Zilberman, N., Yadid, G., Efrati, Y., & Rassovsky, Y. (2020). Who becomes addicted and to what? Psychosocial predictors of substance and behavioral addictive disorders. Psychiatry Research, 291, 113221. https://doi.org/10.1016/j.psychres.2020.113221.
