How do social contexts shape substance use?

Background

Substance use is a growing public health crisis contributing to significant levels of morbidity and mortality (Stockwell et al., 2020; Fischer, 2023). Commonly, substance use has been conceptualized as an individual-level behavioural problem (McNeil, 2021; Dollar et al., 2018; Goodyear, 2021). However, empirical evidence questions the extent to which individual agency shapes observed patterns of substance use (O’Malley, 2009; Gowan et al., 2012). Instead, a growing literature emphasizes the importance of biopsychosocial factors and the interactions across these factors in shaping substance use at the individual and community level (Miller et al., 2013; MacKillop & Ray, 2017; Wallace, 1999; Galea et al., 2004; Kumpfer & Turner, 2009). Given the significant role that social environments play in shaping health and behaviour, it is important to consider the extent to which interpersonal factors shape experiences of substance use. Understanding these relationships can enhance population health and therapeutic strategies to prevent substance use, promote treatment, and reduce harms (Pettersen et al., 2019).

Purpose

The purpose of this evidence brief is to explore the causal pathways linking social environment to substance use patterns and related outcomes. In doing so, we acknowledge that this association has been studied in a variety of ways: with different social and emotional factors having different effects on different outcomes (Copeland et al., 2017, 2018; Kim et al., 2006). As well, we acknowledge that people engage in a wide variety of patterns of substance use and these patterns can be difficult to summarize concisely (Bruno & Csiernik, 2018; Halladay et al., 2020; Clark & Winters, 2002). While such nuances are important to consider, this brief aims to summarize across these by highlighting only the most salient and well-demonstrated findings that are most important to our understanding of the role that the social environment plays in shaping substance use and related outcomes.

Evidence from Existing Studies

Previous research highlights a significant relationship between substance use and the social environments in which it occurs (Stewart et al., 2022; Derrick et al., 2019; Trucco et al., 2020). Multiple pathways contributing to this association (Ingram et al., 2020). Indeed, poor quality social networks, adverse social experiences, loneliness, and social isolation have been identified as strong and consistent predictors for substance use behaviours and related health outcomes (Farmer et al., 2022; Amaro et al., 2021; La Rosa & White, 2001; Gili et al., 2017; Maniglio et al., 2017; Bhuptani et al., 2024; Tretyak et al., 2022; Vrijen et al., 2021; Arcadepani et al., 2021; Klein & Golub et al., 2016; Polenick et al., 2019). Conversely, social capital, social support, belonging, and positive social network factors have been observed to be important contributors to substance use abstention, cessation, recovery, and lower-risk patterns of use (Wills & Vaughan, 1989; Patton et al., 2024; Rudolph et al., 2021; Westmaas et al., 2010; Rathinam & Ezhumalai, 2022; Ikeda et al., 2021; Shahid & Asmat, 2023). These general trends are true across a variety of substances, including both socially accepted drugs (e.g., alcohol; Akerlind & Hornquist, 1992; Calcaterra et al., 2014) and those which are less socially acceptable (e.g., methamphetamine; Ross et al., 2024; Miler et al., 2020).

Case Study: The Rat Park Experiments

In the late 1970s, psychologist Bruce K. Alexander and his team conducted a series of experiments known as the Rat Park studies to explore the impact of environment on drug addiction (Hadaway et al., 1979). Unlike traditional experiments that isolated rats in small cages, Alexander's Rat Park provided a rich, engaging environment where rats lived with access to food, toys, ample space – and most importantly, with other rats(Hadaway et al., 1979). Both the Rat Park and Isolated Rats had access to two types of water: one plain and one laced with morphine. The results were revealing: rats in the socially enriched Rat Park environment overwhelmingly chose plain water, consuming much less morphine than the isolated rats, who consumed high amounts of the drug solution. These findings challenged the notion that drugs alone drive addiction and highlighted the significant role of the social environment, suggesting that poor social conditions are a major factor in substance use and addiction (Mennis et al., 2016). This experiment underscores the importance of a supportive social environment in preventing and treating addiction, advocating for public health strategies that focus on enhancing community and social connections as a means to combat substance use disorders (Alexander et al., 1981).

Explaining the association between social health and substance use related factors, researchers have hypothesized that inter-relationships with mental health are particularly salient (Boateng-Poku et al., 2020; Martin et al., 2023; Prestage et al., 2018; Segrin et al., 2018). Specifically, researchers argue that social distress – and especially early life distress (Kirsch & Lippard, 2022; Kirsch et al., 2020) – may motivate substance use as a means of coping (Cruden & Karmali, 2021; McKirnan et al., 1996). Supporting this, there is a growing body of literature highlighting the involvement of biological mechanisms in driving the association between poor social health and substance use (Pomrenze et al., 2022; Volkow et al., 2011). For example, in studying the connection between opioid addiction and social connectedness, Christie (2021) argues that disruptions to the brain’s opioid system – caused by either poor social conditions or increased substance use – may play an important role in creating vulnerability to addiction. They argue that because the opioid system plays a critical role in reward-processing and social bonding (Panksepp et al., 1980), it is reasonable to assume that opioid use can induce a sense of connection that satisfies our natural motivations to seek out social connection. With these social needs artificially met, individuals may become vulnerable to social isolation as they fail to maintain their social connections to others. In turn, social isolation leads to increased social distress and the cycle begins over again (Sahani et al., 2022). Such processes are empirically supported by evidence showing that opioid administration does indeed effect levels and patterns of sociability (D’Amato & Pavone, 2012); as well as those which emphasize how early life challenges and poor social bonds are strongly correlated with elevated risk for substance use (Ai & Lee, 2018; Lesscher et al., 2015; Sebalo et al., 2023; Martin-Storey et al., 2011; Leza et al., 2021; Hughes et al., 2017; Rogers et al., 2022; Tzouvara et al., 2023; Cortes-Patino et al., 2016) These associations are broadly supported by a vast literature studying how early attachments play a critical role in reward-sensitivity, sociability, and other key psychological and behavioural factors found to underly substance use behaviour (Habibi et al., 2018; Fairbairn et al., 2018). Of course, the extent to which such mechanisms might be involved naturally vary according to the specific substance under consideration because different substances have different interactions with the brain (Quinn et al., 1997). Despite this nuance, available evidence suggests that substance use and related harms are the byproduct of complex interactions between social and reward systems and the social environments that give shape to them.

In addition to these biological and psychological processes, other features of the social environment are observed to contribute to substance use, create difficulties for people who use substances, and worsen substance-use related harms (Conner et al., 2007; Friedman et al., 2020). For example, Allen et al. (2024) showed that rural pregnant women with unmet social and material needs (e.g., food insecurity) have difficulty accessing appropriate services and supports and emphasizes the importance of integrating social supports into prenatal and substance use care. Similarly, using drugs alone and having poorer quality social networks have been linked to increased risk for adverse and preventable harms (Bennett et al., 2022; Falade-Nwulia et al., 2024; Khan et al., 2020). These harms are made worse by the effects of stigma which undermine social support for individuals who may need it most (Matsuyama & Tabuchi, 2024; Krendl & Perry, 2023; Committee on the Science of Changing Behavioral Health Social Norms, 2016; Dyal & Valente, 2015). For example, Wootton et al. (2021) show that individuals who initiate tobacco often experience increases in loneliness – perhaps because the stigma surrounding tobacco use drives social exclusion and withdrawl. In turn, people who use substances experience deleterious health outcomes, including difficulty accessing substance use treatment and harm reduction services (Douglass et al., 2023; Zwick et al., 2020; Yang et al., 2018; Earnshaw, 2021).

Complicating the association between social factors and substance use, it is also important to note that substance use is often used explicitly as part of one’s social interactions with others (Cole et al., 2024; Votaw & Witkiewitz, 2021; Strickland & Smith, 2015). As such, substance use can sometimes be linked to positive dimensions of social health. For example, Bragard et al. (2022) document that on days when individuals felt either particularly low or particularly high levels of loneliness they drank more alcohol. Similarly, higher loneliness was actually related to lower levels of cannabis use. Other studies have similarly shown that daily experiences of connection can correlate with daily patterns of substance use (Parnes et al., 2024). These findings likely reflect the reality that alcohol and cannabis use are often used to facilitate social behaviour (Goldstick et al., 2022). Further supporting this, studies examining the effect of the COVID-19 lockdowns on substance use highlight shifting substance use, such as decreased cocaine use (another drug frequently used in social contexts; Christie et al., 2021).

In a similar vein, substance use can facilitate belonging and inclusion by helping individuals “fit in” with their particular social group (Heijden et al., 2024; Bourne & Weatherburn, 2017; Zhang et al., 2022; Eisenberg et al., 2014). This is particularly true when social norms are permissive of substance use (Stoddard et al., 2012; Mehanovic et al., 2022). Of course, some substances are socially acceptable while others are not (Bolinski et al., 2019; Pilatti et al., 2021; Melander et al., 2016). Moreover, some substances are socially acceptable in some subgroups (or by some parts of one’s social network; Marziali et al., 2022; Oetting et al., 1998; Burgess et al., 2024). As such, different substances may have different relationships with social factors, including norms-related factors. Furthermore, it is important to recognize that individuals may have inaccurate perceptions of social norms, leading to greater risk for substance use among those who perceive norms to be more accepting than they are (Amialchuk et al., 2019; Carter & Kahnweiler, 2000). Regardless of perceptions, normative patterns and social influences such as those described above create different patterns of substance use across populations and are a key mechanism linking social environments to substance use and related harms (Studer et al., 2016; Steinmetz-Wood et al., 2018; Dempsey et al., 2018). Supporting this, Peviani et al. (2020) and others provide empirical evidence that parent and peer substance use (and attitudes about substance use) play important roles in shaping substance use behaviour (Meulewaeter et al., 2022; Averna & Hesselbrock, 2001; Barnow et al., 2002; Wasserman et al., 2001; Rudolph et al., 2020; Falkin & Strauss, 2003).

On the other hand, our relationships can reinforce protective elements and provide positive outlets for coping (Wills & Cleary, 1996; Arslan, 2022). In fact, beyond their normative effects, social networks also play important roles in shaping substance use and related outcomes by providing social support that may buffer against the stressors that lead to substance use (Black & Chung, 2014; Kahn et al., 2022; Sharam et al., 2016; Moos et al., 2007; Cummings et al., 2022; Rogers et al., 2023). Supporting this, Boumans et al. (2022) identify social contact from supportive peers and family members as a key ingredient of effective recovery programs (Hogue et al., 2021; Kympfer et al., 2003; ). Social support can also help individuals develop self-esteem and self-efficacy, which in turn are helpful in managing substance use and coping (Cao & Liang, 2017; Scheer et al., 2022; Khatib et al., 2012; Yang et al., 2019; Lau et al., 2023; Homayuni & Hosseini, 2023). This is exemplified by studies such as that by Stevens et al. (2015), which showed how social support and social network features are linked to abstinence self-efficacy among individuals living in recovery houses and participating in 12-step support groups (Donovan et al., 2013).

Another important social factor to consider is loneliness and its relationship to both substance use and social behaviour. Loneliness represents the subjective evaluations about one’s social environment and wellbeing (Perlman & Peplau, 1984). Furthermore, loneliness is found to co-occur along a variety of maladaptive social cognitions, such as social distrust, vigilance, self-preservation, and fear of negative evaluations (Hawkley & Cacioppo, 2013; Pelot et al., 2020; Moriarity et al., 2021). It is likely that such cognitive schema contribute, in part, to adverse personal and social outcomes related to substance use. For example, Ingram and colleagues (2020) speak to how some of the cognitive features of loneliness can lead to unhelpful interpersonal behaviours that undermine the development of high quality relationships (Dingle et al., 2018; Unger et al., 2003; Harrison et al., 2017). Thus, it is reasonable to conceptualize the intersection of substance use and loneliness as a “syndemic” in which the co-occurrence of these conditions synergistically work together to the disadvantage of those who are affected (Mesias-Gazmuri et al., 2023). As such, it may be beneficial (and perhaps even necessary) to address social environments in order to reduce substance use (Ingram et al., 2020; de Espindola et al., 2020; Majer et al., 2016; Elkhalifa et al., 2020; Kumar et al., 2016; Pettersen et al., 2019). Fortunately, group-based supports are already widely used (Lopez, 2021).

Analyses By the Canadian Alliance for Social Connection and Health

Using data from the Canadian Social Connection Survey (n = 1600), we investigated the relationship between loneliness and the use of various substances, including alcohol, tobacco, and other drugs. Data were collected on substance use frequency (i.e., 7-point scale “Not in the past six months” to “Daily or almost daily”) and loneliness scores using the De Jong Gierveld Emotional and Social Loneliness Scale. To investigate whether loneliness was related to patterns of substance use, zero-inflated Poisson regression models were employed. These models allowed us to examine both the frequency of substance use (count model) and the likelihood of being a non-user (zero-inflation model).

As shown in Figure 1, most participants reported abstaining from most substances and there were similar distributions in the frequency of use.

Alcohol. Higher levels of loneliness were associated with a slight decrease in the frequency of alcohol use (Estimate = -0.034, SE = 0.010, z = -3.32, p < .001). Additionally, lonelier individuals were somewhat less likely to completely abstain from alcohol (Estimate = -0.094, SE = 0.040, z = -2.36, p = .018).

Tobacco. Lonelier individuals tended to use tobacco less frequently (Estimate = -0.050, SE = 0.013, z = -3.75, p < .001). Furthermore, those with higher loneliness scores were significantly less likely to be non-users of tobacco (Estimate = -0.189, SE = 0.031, z = -6.14, p < .001).

Nicotine (Other than Tobacco). There was a slight, non-significant decrease in nicotine use with higher loneliness (Estimate = -0.027, SE = 0.020, z = -1.36, p = .174). However, lonelier individuals were much less likely to avoid nicotine entirely (Estimate = -0.259, SE = 0.034, z = -7.69, p < .001).

Cannabis. Higher loneliness slightly reduced the frequency of cannabis use, though the decrease was marginal (Estimate = -0.030, SE = 0.016, z = -1.84, p = .066). Loneliness greatly reduced the likelihood of being a non-user of cannabis (Estimate = -0.275, SE = 0.033, z = -8.31, p < .001).

Hallucinogens. Lonelier people showed a non-significant increase in the use of hallucinogens (Estimate = 0.042, SE = 0.031, z = 1.37, p = .172). However, loneliness significantly reduced the chances of being in the group that never uses hallucinogens (Estimate = -0.308, SE = 0.040, z = -7.69, p < .001).

Methamphetamine. Loneliness was associated with a slight, non-significant increase in methamphetamine use (Estimate = 0.020, SE = 0.031, z = 0.63, p = .526). Lonelier individuals were significantly less likely to be non-users of methamphetamine (Estimate = -0.333, SE = 0.040, z = -8.24, p < .001).

Cocaine. There was a non-significant increase in the frequency of cocaine use with higher loneliness (Estimate = 0.017, SE = 0.033, z = 0.53, p = .594). However, loneliness substantially decreased the probability of completely avoiding cocaine (Estimate = -0.319, SE = 0.041, z = -7.83, p < .001).

Inhalants. Loneliness showed a very slight, non-significant increase in inhalant use (Estimate = 0.008, SE = 0.034, z = 0.24, p = .810). Lonelier individuals were significantly less likely to be non-users of inhalants (Estimate = -0.325, SE = 0.041, z = -7.96, p < .001).

Heroin. There was a non-significant increase in heroin use among lonelier people (Estimate = 0.029, SE = 0.033, z = 0.88, p = .382). Loneliness greatly reduced the likelihood of abstaining from heroin (Estimate = -0.300, SE = 0.041, z = -7.35, p < .001).

Painkillers (Non-prescribed). Loneliness showed a non-significant increase in the use of non-prescribed painkillers (Estimate = 0.038, SE = 0.030, z = 1.26, p = .206). Those who are lonelier were significantly less likely to never use non-prescribed painkillers (Estimate = -0.292, SE = 0.040, z = -7.22, p < .001).

Sedatives. Loneliness was associated with a non-significant increase in the use of sedatives (Estimate = 0.019, SE = 0.027, z = 0.71, p = .480). Lonelier individuals were much less likely to be non-users of sedatives (Estimate = -0.324, SE = 0.039, z = -8.39, p < .001).

Stimulants. There was no significant association between loneliness and the frequency of stimulant use (Estimate = -0.010, SE = 0.029, z = -0.33, p = .740). However, loneliness significantly reduced the likelihood of completely avoiding stimulants (Estimate = -0.323, SE = 0.039, z = -8.24, p < .001).

Over-the-Counter Drugs (Non-prescribed). Loneliness showed a non-significant increase in the use of non-prescribed over-the-counter drugs (Estimate = 0.012, SE = 0.020, z = 0.59, p = .556). Lonelier individuals were significantly less likely to be in the group that never uses over-the-counter drugs without a prescription (Estimate = -0.234, SE = 0.035, z = -6.65, p < .001).

As demonstrated by the findings above, we found that higher levels of loneliness increases the risk of substance use across various drugs, suggesting that lonely individuals might turn to substances as a coping mechanism. However, for substances typically used in social contexts, such as alcohol, tobacco, and cannabis, loneliness was associated with decreased frequency of use. This pattern might imply that lonelier individuals might have fewer opportunities for social substance use, thus using these substances less frequently. Of course, further investigation is needed in samples with a higher proportion of substance use and more varied frequency. Nevertheless, despite inherent limitations in sampling and data analyses, these findings support the literature reviewed above which emphasize links between loneliness and substance use.

Discussion

The evidence reviewed highlights the multifaceted role of social environments in influencing substance use and addiction. In doing so, we highlight both risk and protective factors. Early life adversities, such as poor attachment styles and emotional regulation challenges, set a foundation where substances may be used as coping mechanisms. The stigma and social rejection associated with drug use exacerbate these issues, making recovery and social reintegration challenging, and often leading to relapse. Conversely, supportive social environments that set positive norms and facilitate and community engagement are crucial in enabling access to harm reduction and recovery services, which can significantly mitigate the adverse effects of substance use. These sources of support build self-esteem and self-efficacy and provide the emotional and tangible supports necessary to overcome the biological dependence associated with many substances. Of course, the relationships reviewed are complex and require continued exploration with more comprehensive datasets, collected from larger samples over longer periods of time. As well, network and dyadic data can improve our ability to understand the underlying social processes that are described. Nevertheless, despite the needs for increasing methodological rigor in the existing literature, it is apparent that there is much to be gained by prioritizing the development and maintenance of healthy social relationships in order to reduce unwanted substance use or substance use related harms (Strickland & Acuff, 2023). It is unlikely that considerable population-wide success in treatment and prevention will be achieved without addressing the social conditions that give rise to and sustain difficulties associated with substance use.

Conclusion

Based on the available evidence, we recommend that communities and public health initiatives focus on enhancing social support systems and reducing stigma to better address substance use and support recovery. Investments are needed to promote the social health of people who use substances as well as the broader population in order to protect them from potentially adverse harms of substance use.