Stress is a commonly reported precipitant to drinking; however, research studies only show an inconsistent effect of stress on alcohol consumption. This study addresses previous experimental study limitations in a rigorous test of the tension reduction hypothesis.
l
Drinking to cope with aversive experiences or emotions is commonly reported among people who experience harms from alcohol. This notion, that people consume alcohol to alleviate negative feelings, is central to hypotheses attempting to explain addiction, such as the tension reduction hypothesis, which states that alcohol is consumed in part as an attempt to reduce stress or tension, and therefore that alcohol use should be more likely in stressful, compared to non-stressful, moments. Although laboratory-based studies have found consistent support for this tendency for people to consume alcohol to relieve negative affect, real world naturalistic studies using ecological momentary assessment, in which people take multiple surveys throughout the day, have found less consistent support, with a recent meta-analysis suggesting that same-day positive affect, not negative affect/stress or tension, predicts same day increases in alcohol consumption. The laboratory based experimental studies often do not include non-alcoholic alternatives (e.g., coffee, soda, etc.), an experimental arrangement which does not reflect the real world and may explain the greater-than-expected support for the tension reduction hypothesis. Further, most studies have explored effects of stress on the amount of alcohol consumed rather than the decision to initiate alcohol consumption – different processes may explain the decision to initiate alcohol use versus the decision to continue alcohol use. The current laboratory-based experimental study tested whether stress leads individuals to initiate alcohol consumption versus initiating consumption of a an appealing non-alcoholic alternatives and to better understand the decision-making processes that underlie these choices.
The researchers used a 2 x 2 factorial experimental design to test whether alcohol exposure (alcohol versus no alcohol conditions) and acute stress (stress and no stress conditions) impacted decisions to consume alcohol and non-alcoholic beverages. The researchers recruited and randomized 160 adults with no history of alcohol use disorder treatment. Participants were 31 years old, on average (SD = 7.6); 56% were male and 44% were female. Almost half (46%) were White, 15% were Asian, 14% Hispanic/Latino, 13% were Black, and 12% were multiracial. The median alcohol use disorder identification score was 8, consistent with risky alcohol use, but not necessarily alcohol use disorder.
First, participants rated 30 alcoholic and 30 non-alcoholic beverages based on preference. Next, all participants completed a forced choice task in which participants made different choices between alcoholic and non-alcohol beverages. This task created two of the primary outcomes: proportion of alcohol, relative to non-alcoholic, beverage choice, and decision speed. Next, depending on the which condition the participant was randomized to, the participants would either consume alcohol (until their breath alcohol concentration reached 0.06%) or non-alcohol beverages in a simulated bar. Then they would either listen to a neutral recording or to a recording of a personal stressful memory (which they provided prior to the experimental session). After the manipulation, all participants once again completed the forced choice task to determine whether alcohol choice and decision speed changed due to the alcohol and stress manipulations. The researchers wanted to determine whether participants shifted preference toward alcohol choices in the forced choice paradigm, chose alcohol faster, and purchased more drinks in the alcohol purchase task, after the manipulation.
First, the researchers checked whether the stress and alcohol inductions worked properly. As expected, participants in the alcohol condition reported greater alcohol effects relative to those who did not receive alcohol, and participants in a stress condition reported greater overall stress, and greater decreases in mood, compared to those who just listened to a neutral recording.
Next, the researchers examined differences in alcohol choice following the manipulations, after controlling for pre-alcohol or stress alcohol choice. Those in the stress manipulation were more likely to select alcohol relative to nonalcoholic beverages; however, there were no differences in alcohol choice between those who consumed alcohol and those who did not, suggesting that people were about equally likely to choose alcohol regardless of whether they had not consumed alcohol or were at a BAC of 0.06%. This suggests a potential interaction effect in which stress might be more influential on the decision to consume alcohol when people have not yet started to drink. These results were even stronger when the researchers removed participants who did not respond to the stress induction to test the robustness of the findings. Although there were differences in alcohol choice, there were no differences in reaction time.
The researchers also observed results on mechanisms which suggested that, when sober, stress may result in choosing alcohol despite preferring a non-alcoholic option.
These results support the tension reduction hypothesis for this group of younger individuals who typically engaged in risky levels of alcohol use, on average, and suggest that stress may lead to increased preference for alcohol, at least before alcohol consumption has begun. Continued drinking after beginning to drink is likely driven by a different factor other than stress – for example, more elated feelings (euphoria) or feeling psychologically looser and freer (disinhibited) after consuming alcohol may reinforce continued use. The current study deviated from previous studies testing tension reduction hypothesis with alcohol alone by including non-alcoholic beverages that may compete with alcohol, to more closely emulate the real world. Though studies like these are a step in the right direction, non-alcoholic beverages may not serve the same behavioral function as alcoholic beverages in stressful situations and therefore are likely not true “alternatives” to coping with stress. Further, while stress is an important emotional state that many people identify as precipitating alcohol consumption, other emotional states such as sadness, anxiety, or fear may demonstrate different effects on alcohol preference. Also, the researchers did not conduct a test of inducing positive affect that may have served as a useful test of the real-world findings that people drink in the real-world when they’re feeling good, not bad. In the same way, participants might be encouraged to think about their most exciting or happiest memories or a future event they’re looking forward to, in order to examine the effect of positive emotion/excitement on alcohol use compared to negative affect/stress.
Even so, these data provided a more stringent experimental test of the tension reduction hypothesis providing more confidence in its utility, at least for beginning to drink alcohol rather than continue to drink. This does not help to resolve the disparity in findings between experimental and naturalistic studies that often find that same day negative emotions do not lead to drinking. This may be for many reasons, such as a greater number of confounding factors in the real world and issues with measurement or that participants in laboratory-based studies may feel they are supposed to drink when experiencing stress versus not which then leads to alcohol consumption. More work is needed to better understand these differences between lab-based and real-world studies.
Stress may lead to a greater alcohol preference when choosing between alcohol and non-alcoholic beverages, at least before people start drinking.
Dora, J., Shinn, M., Copeland, A., Neilson, E. C., Weiss, N. H., Witkiewitz, K., Murphy, J. G., Field, M., George, W. H., & King, K. M. (2026). How people decide to consume (more) alcohol when feeling stressed. Addiction, 121(2), 370-387. doi: 10.1111/add.70213.
l
Drinking to cope with aversive experiences or emotions is commonly reported among people who experience harms from alcohol. This notion, that people consume alcohol to alleviate negative feelings, is central to hypotheses attempting to explain addiction, such as the tension reduction hypothesis, which states that alcohol is consumed in part as an attempt to reduce stress or tension, and therefore that alcohol use should be more likely in stressful, compared to non-stressful, moments. Although laboratory-based studies have found consistent support for this tendency for people to consume alcohol to relieve negative affect, real world naturalistic studies using ecological momentary assessment, in which people take multiple surveys throughout the day, have found less consistent support, with a recent meta-analysis suggesting that same-day positive affect, not negative affect/stress or tension, predicts same day increases in alcohol consumption. The laboratory based experimental studies often do not include non-alcoholic alternatives (e.g., coffee, soda, etc.), an experimental arrangement which does not reflect the real world and may explain the greater-than-expected support for the tension reduction hypothesis. Further, most studies have explored effects of stress on the amount of alcohol consumed rather than the decision to initiate alcohol consumption – different processes may explain the decision to initiate alcohol use versus the decision to continue alcohol use. The current laboratory-based experimental study tested whether stress leads individuals to initiate alcohol consumption versus initiating consumption of a an appealing non-alcoholic alternatives and to better understand the decision-making processes that underlie these choices.
The researchers used a 2 x 2 factorial experimental design to test whether alcohol exposure (alcohol versus no alcohol conditions) and acute stress (stress and no stress conditions) impacted decisions to consume alcohol and non-alcoholic beverages. The researchers recruited and randomized 160 adults with no history of alcohol use disorder treatment. Participants were 31 years old, on average (SD = 7.6); 56% were male and 44% were female. Almost half (46%) were White, 15% were Asian, 14% Hispanic/Latino, 13% were Black, and 12% were multiracial. The median alcohol use disorder identification score was 8, consistent with risky alcohol use, but not necessarily alcohol use disorder.
First, participants rated 30 alcoholic and 30 non-alcoholic beverages based on preference. Next, all participants completed a forced choice task in which participants made different choices between alcoholic and non-alcohol beverages. This task created two of the primary outcomes: proportion of alcohol, relative to non-alcoholic, beverage choice, and decision speed. Next, depending on the which condition the participant was randomized to, the participants would either consume alcohol (until their breath alcohol concentration reached 0.06%) or non-alcohol beverages in a simulated bar. Then they would either listen to a neutral recording or to a recording of a personal stressful memory (which they provided prior to the experimental session). After the manipulation, all participants once again completed the forced choice task to determine whether alcohol choice and decision speed changed due to the alcohol and stress manipulations. The researchers wanted to determine whether participants shifted preference toward alcohol choices in the forced choice paradigm, chose alcohol faster, and purchased more drinks in the alcohol purchase task, after the manipulation.
First, the researchers checked whether the stress and alcohol inductions worked properly. As expected, participants in the alcohol condition reported greater alcohol effects relative to those who did not receive alcohol, and participants in a stress condition reported greater overall stress, and greater decreases in mood, compared to those who just listened to a neutral recording.
Next, the researchers examined differences in alcohol choice following the manipulations, after controlling for pre-alcohol or stress alcohol choice. Those in the stress manipulation were more likely to select alcohol relative to nonalcoholic beverages; however, there were no differences in alcohol choice between those who consumed alcohol and those who did not, suggesting that people were about equally likely to choose alcohol regardless of whether they had not consumed alcohol or were at a BAC of 0.06%. This suggests a potential interaction effect in which stress might be more influential on the decision to consume alcohol when people have not yet started to drink. These results were even stronger when the researchers removed participants who did not respond to the stress induction to test the robustness of the findings. Although there were differences in alcohol choice, there were no differences in reaction time.
The researchers also observed results on mechanisms which suggested that, when sober, stress may result in choosing alcohol despite preferring a non-alcoholic option.
These results support the tension reduction hypothesis for this group of younger individuals who typically engaged in risky levels of alcohol use, on average, and suggest that stress may lead to increased preference for alcohol, at least before alcohol consumption has begun. Continued drinking after beginning to drink is likely driven by a different factor other than stress – for example, more elated feelings (euphoria) or feeling psychologically looser and freer (disinhibited) after consuming alcohol may reinforce continued use. The current study deviated from previous studies testing tension reduction hypothesis with alcohol alone by including non-alcoholic beverages that may compete with alcohol, to more closely emulate the real world. Though studies like these are a step in the right direction, non-alcoholic beverages may not serve the same behavioral function as alcoholic beverages in stressful situations and therefore are likely not true “alternatives” to coping with stress. Further, while stress is an important emotional state that many people identify as precipitating alcohol consumption, other emotional states such as sadness, anxiety, or fear may demonstrate different effects on alcohol preference. Also, the researchers did not conduct a test of inducing positive affect that may have served as a useful test of the real-world findings that people drink in the real-world when they’re feeling good, not bad. In the same way, participants might be encouraged to think about their most exciting or happiest memories or a future event they’re looking forward to, in order to examine the effect of positive emotion/excitement on alcohol use compared to negative affect/stress.
Even so, these data provided a more stringent experimental test of the tension reduction hypothesis providing more confidence in its utility, at least for beginning to drink alcohol rather than continue to drink. This does not help to resolve the disparity in findings between experimental and naturalistic studies that often find that same day negative emotions do not lead to drinking. This may be for many reasons, such as a greater number of confounding factors in the real world and issues with measurement or that participants in laboratory-based studies may feel they are supposed to drink when experiencing stress versus not which then leads to alcohol consumption. More work is needed to better understand these differences between lab-based and real-world studies.
Stress may lead to a greater alcohol preference when choosing between alcohol and non-alcoholic beverages, at least before people start drinking.
Dora, J., Shinn, M., Copeland, A., Neilson, E. C., Weiss, N. H., Witkiewitz, K., Murphy, J. G., Field, M., George, W. H., & King, K. M. (2026). How people decide to consume (more) alcohol when feeling stressed. Addiction, 121(2), 370-387. doi: 10.1111/add.70213.
l
Drinking to cope with aversive experiences or emotions is commonly reported among people who experience harms from alcohol. This notion, that people consume alcohol to alleviate negative feelings, is central to hypotheses attempting to explain addiction, such as the tension reduction hypothesis, which states that alcohol is consumed in part as an attempt to reduce stress or tension, and therefore that alcohol use should be more likely in stressful, compared to non-stressful, moments. Although laboratory-based studies have found consistent support for this tendency for people to consume alcohol to relieve negative affect, real world naturalistic studies using ecological momentary assessment, in which people take multiple surveys throughout the day, have found less consistent support, with a recent meta-analysis suggesting that same-day positive affect, not negative affect/stress or tension, predicts same day increases in alcohol consumption. The laboratory based experimental studies often do not include non-alcoholic alternatives (e.g., coffee, soda, etc.), an experimental arrangement which does not reflect the real world and may explain the greater-than-expected support for the tension reduction hypothesis. Further, most studies have explored effects of stress on the amount of alcohol consumed rather than the decision to initiate alcohol consumption – different processes may explain the decision to initiate alcohol use versus the decision to continue alcohol use. The current laboratory-based experimental study tested whether stress leads individuals to initiate alcohol consumption versus initiating consumption of a an appealing non-alcoholic alternatives and to better understand the decision-making processes that underlie these choices.
The researchers used a 2 x 2 factorial experimental design to test whether alcohol exposure (alcohol versus no alcohol conditions) and acute stress (stress and no stress conditions) impacted decisions to consume alcohol and non-alcoholic beverages. The researchers recruited and randomized 160 adults with no history of alcohol use disorder treatment. Participants were 31 years old, on average (SD = 7.6); 56% were male and 44% were female. Almost half (46%) were White, 15% were Asian, 14% Hispanic/Latino, 13% were Black, and 12% were multiracial. The median alcohol use disorder identification score was 8, consistent with risky alcohol use, but not necessarily alcohol use disorder.
First, participants rated 30 alcoholic and 30 non-alcoholic beverages based on preference. Next, all participants completed a forced choice task in which participants made different choices between alcoholic and non-alcohol beverages. This task created two of the primary outcomes: proportion of alcohol, relative to non-alcoholic, beverage choice, and decision speed. Next, depending on the which condition the participant was randomized to, the participants would either consume alcohol (until their breath alcohol concentration reached 0.06%) or non-alcohol beverages in a simulated bar. Then they would either listen to a neutral recording or to a recording of a personal stressful memory (which they provided prior to the experimental session). After the manipulation, all participants once again completed the forced choice task to determine whether alcohol choice and decision speed changed due to the alcohol and stress manipulations. The researchers wanted to determine whether participants shifted preference toward alcohol choices in the forced choice paradigm, chose alcohol faster, and purchased more drinks in the alcohol purchase task, after the manipulation.
First, the researchers checked whether the stress and alcohol inductions worked properly. As expected, participants in the alcohol condition reported greater alcohol effects relative to those who did not receive alcohol, and participants in a stress condition reported greater overall stress, and greater decreases in mood, compared to those who just listened to a neutral recording.
Next, the researchers examined differences in alcohol choice following the manipulations, after controlling for pre-alcohol or stress alcohol choice. Those in the stress manipulation were more likely to select alcohol relative to nonalcoholic beverages; however, there were no differences in alcohol choice between those who consumed alcohol and those who did not, suggesting that people were about equally likely to choose alcohol regardless of whether they had not consumed alcohol or were at a BAC of 0.06%. This suggests a potential interaction effect in which stress might be more influential on the decision to consume alcohol when people have not yet started to drink. These results were even stronger when the researchers removed participants who did not respond to the stress induction to test the robustness of the findings. Although there were differences in alcohol choice, there were no differences in reaction time.
The researchers also observed results on mechanisms which suggested that, when sober, stress may result in choosing alcohol despite preferring a non-alcoholic option.
These results support the tension reduction hypothesis for this group of younger individuals who typically engaged in risky levels of alcohol use, on average, and suggest that stress may lead to increased preference for alcohol, at least before alcohol consumption has begun. Continued drinking after beginning to drink is likely driven by a different factor other than stress – for example, more elated feelings (euphoria) or feeling psychologically looser and freer (disinhibited) after consuming alcohol may reinforce continued use. The current study deviated from previous studies testing tension reduction hypothesis with alcohol alone by including non-alcoholic beverages that may compete with alcohol, to more closely emulate the real world. Though studies like these are a step in the right direction, non-alcoholic beverages may not serve the same behavioral function as alcoholic beverages in stressful situations and therefore are likely not true “alternatives” to coping with stress. Further, while stress is an important emotional state that many people identify as precipitating alcohol consumption, other emotional states such as sadness, anxiety, or fear may demonstrate different effects on alcohol preference. Also, the researchers did not conduct a test of inducing positive affect that may have served as a useful test of the real-world findings that people drink in the real-world when they’re feeling good, not bad. In the same way, participants might be encouraged to think about their most exciting or happiest memories or a future event they’re looking forward to, in order to examine the effect of positive emotion/excitement on alcohol use compared to negative affect/stress.
Even so, these data provided a more stringent experimental test of the tension reduction hypothesis providing more confidence in its utility, at least for beginning to drink alcohol rather than continue to drink. This does not help to resolve the disparity in findings between experimental and naturalistic studies that often find that same day negative emotions do not lead to drinking. This may be for many reasons, such as a greater number of confounding factors in the real world and issues with measurement or that participants in laboratory-based studies may feel they are supposed to drink when experiencing stress versus not which then leads to alcohol consumption. More work is needed to better understand these differences between lab-based and real-world studies.
Stress may lead to a greater alcohol preference when choosing between alcohol and non-alcoholic beverages, at least before people start drinking.
Dora, J., Shinn, M., Copeland, A., Neilson, E. C., Weiss, N. H., Witkiewitz, K., Murphy, J. G., Field, M., George, W. H., & King, K. M. (2026). How people decide to consume (more) alcohol when feeling stressed. Addiction, 121(2), 370-387. doi: 10.1111/add.70213.
SIGN UP FOR THE BULLETIN
151 Merrimac St., 4th Floor. Boston, MA 02114
