Abstract
Background:
Cognitive functioning refers to storage and manipulation of information and includes executive functioning (EF) and attention (ATT). While physical activity (PA) improves cognitive functioning, decrements are associated with frequent substance use. This study examined PA on cognitive functioning within the context of past-year substance use.
Methods:
Using NESARC–III data (N=36,309), cross-sectional analysis examined interactions between self-reported past-year PA and substance use in relation to cognitive functioning.
Results:
As hypothesized, light-to-moderate, vigorous, and total PA conditional main effects were positively associated with both facets of cognition, while frequent substance use conditional main effects were negatively associated with ATT and EF. The positive association between PA and cognition was diminished by substance use. Frequent binge drinking, marijuana, cocaine, and opioid use weakened the impact of light-to-moderate PA on EF, and only frequent cocaine use lessened the relationship between vigorous PA on EF. When PA intensities were combined, frequent binge drinking and cocaine use weakened the PA and EF association. Infrequent stimulant use reduced the association between all levels of PA and ATT, while infrequent marijuana use unexpectedly enhanced the relation between vigorous PA and ATT.
Conclusions:
Overall, PA enhanced two facets of cognitive functioning across six substances. However, these benefits are reduced in the context of frequency of substance use. The positive association between light-to-moderate PA and EF appears to be more sensitive in the context of frequent substance use than vigorous PA. Implications for public health messaging and PA as cognitive remediation treatment for substance use disorders are discussed.
Keywords: executive functioning, attention, alcohol, opioids, marijuana, exercise
1. INTRODUCTION
Executive functioning (EF) and attention (ATT), broadly defined as aspects of cognition, are malleable and are affected positively or negatively by engagement in health behaviors such as substance use and physical activity (PA; Aharonovich et al., 2017; Audiffren and André, 2019). Both EF (e.g., planning, considering consequences) and ATT (e.g., multitasking, sustained attention) are important for engaging in goal-directed behavior and daily functioning (Diamond, 2013). Chronic heavy substance use is detrimental to EF and ATT (Hagen et al., 2016; Ramey and Regier, 2019), and individuals with pre-morbid cognitive impairments are more likely to use substances regularly (Morin et al., 2018). Meanwhile, PA broadly improves cognition (Moreau and Chou, 2019), with specific yet non-uniform positive effects noted for EF and other components of cognition (Erickson et al., 2019); yet, how these behaviors interact and impact cognition overall is unknown. The presence of regular PA may mitigate or offset the negative effects of chronic heavy substance use (Costa et al., 2019). The aim of this paper was therefore to examine EF and ATT in relation to regular PA and frequent substance use in a large epidemiological sample.
Different substances (e.g., alcohol, marijuana, cocaine, opioids) have differential negative effects on cognition broadly and its respective components (e.g., EF, working memory) depending upon the drug’s biological mechanism of action, frequency and intensity of use, and these effects can persist even after periods of abstinence (Hagen et al., 2016; Manning et al., 2017; Mulhauser et al., 2018; Ramey and Regier, 2019). For instance, Carbia and colleagues (2018) in a longitudinal study found that young adults who binge drank over time performed significantly worse on facets of EF (e.g., weaker inhibitory control, poorer working memory) compared to non-binge drinkers. Deficits in ATT and EF are also seen in individuals who chronically use opioids as compared to healthy controls (e.g., Allegri et al., 2019; Wollman et al., 2019). Cognitive deficits are also known to vary widely depending on many other individual difference factors (e.g., genetic predisposition, sex, level of education, socioeconomic status; c.f., Le Berre et al., 2017). Yet, the trend is consistent that greater frequency and intensity of substance use is associated with greater cognitive impairment, which is associated with impaired daily functioning and a host of negative social, occupational, and health outcomes (Sampedro-Piquero et al., 2019). While cognition is impaired by frequent substance use, the converse is true as well. Developmental studies find impairments in cognition can lead to greater likelihood of initiating and regularly using substances (e.g., Nigg et al., 2004).
Conversely, PA - both acute and chronic - positively affects cognition (Audiffren and André, 2019; Erickson et al., 2019; Kara et al., 2013; Krell-Roesch et al., 2019; Moreau and Chou, 2019; Opel et al., 2019). For example, a meta-analysis of 28 studies found a small effect (d=0.34) on EF for acute vigorous intensity PA when compared to a rest condition. Chronic exercise (i.e., moderate to vigorous PA) preserves cognition in individuals with greater lifetime stress, and clinical trials of PA show improvement in various aspects of EF over time (e.g., cognitive flexibility, working memory, response inhibition; Audiffren and André, 2019; Head et al., 2012; Masley et al., 2009). Thus, PA is particularly beneficial in individuals at high-risk for cognitive impairments. Yet, a US government expert review panel that examined the impact of PA on cognition across the life span concluded insufficient evidence exists to determine the effects of moderate to vigorous PA in the general adult population (Erickson et al., 2019). The current study will contribute substantially to this scant literature.
To date, few studies have examined how substance use and PA interact in their effect on cognition. Costa and colleagues (2019), in a review of exercise as an intervention for substance use disorders (SUD), concluded that exercise may improve response inhibition, which frequently is impaired in SUD populations. Animal models of SUD also find that exercise can reduce drug seeking behavior (Galaj et al., 2020) with one hypothesized mechanism of action being normalization of abnormal dopaminergic functioning (Costa et al., 2019). However, no studies to our knowledge have examined the interaction of PA and substance use on cognition in a general population sample.
The National Epidemiologic Survey on Alcohol and Related Conditions-III (NESARC-III) is a nationally representative survey on adults (≥18 years) living in the United States and contains self-report measures of substance use, PA (frequency and duration), and cognition. Using NESARC-III, Aharonovich and colleagues (2017) found that ATT and EF were significantly lower in individuals who frequently used substances compared to less frequent users and abstainers. This study will extend Aharonvich et al.’s (2017) findings by incorporating PA into the analyses. We hypothesize that PA will have a small but positive effect on cognition and that it will mitigate, but not eliminate, the negative effects of frequent substance use on cognition.
2. METHODS
2.1. Design
A cross-sectional analysis using psychometrically-supported measures of substance use, cognition, and PA. Human subjects approval was granted by the Saint Louis University Institutional Review Board.
2.2. Data
The NESARC-III dataset (N = 36,309) was used for the present study (Grant et al., 2015a; Grant et al., 2015b), which involved multi-stage probability sampling to create a representative sample of non-institutionalized US adults (age ≥18) who were not active duty military. Informed consent was obtained, assessments were conducted in-person using computer-assisted personal interviewing, and participants were compensated $90. Overall survey response rate was 60.1%. Some participants (n=761) did not respond to the questions of interest. The final sample size used in this study was 35,548 participants. Data were subsequently adjusted for non-response and weighted to represent the adult US general population.
2.3. Measures
Demographics.
Biological sex was a dichotomous variable. Age was categorized into 18 to 24 years, 25 to 44 years, 45 to 64 years, 65 to 74 years, and 75 years or older. Race was classified into non-Hispanic Whites, African Americans, American Indian/Alaska Native, Asian/Native Hawaiian/Other Pacific Islander, and Hispanic. Education was categorized as less than high school, high school graduates or GED, some college, a college degree, and graduate or higher. Personal income was classified into less than $20,000, $20,000 to $34,999, $35,000 to $69,999, and $70,000 or higher.
Past-year substance use.
The Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (Grant et al., 2015b) assessed frequency of substance use, including binge drinking, marijuana, cocaine, opioids, sedatives, and non-medical use of stimulants. Frequency of use was subsequently coded into a three-level variable (i.e., no use, infrequent use, frequent use). Frequent use was defined as use from once per week to daily. Infrequent use was defined as 1–2 times per year to 2–3 times per month. No use represented individuals who did not report using the substance in the past year.
Physical Activity.
Four questions assessed frequency and intensity of PA in the past year based upon the International Physical Activity Questionnaire – Short Form (Lee et al., 2011). Participants were asked about how often in the past 12 months they engaged in (1) vigorous and (2) light-to-moderate physical activities. Response options for these two items were on an 11-point Likert scale and ranged from never to every day. Individuals reporting any engagement in these activities were asked how long (minutes) they usually did these activities each time. From this set of questions two variables were subsequently calculated: (1) vigorous calories expended per day and (2) light-to-moderate calories expended per day. The calories per day are a measure of overall PA volume. Formulas for converting frequency and intensity (minutes) into calories followed American College of Sports Medicine guidelines (American College of Sports Medicine, 2017), were based upon sex, height, weight, and used metabolic equivalents values of 8 and 3. A third variable was also created, total PA, summing light-to-moderate PA with vigorous PA. The rationale for this delineation is that the benefits of light versus other intensities of PA are commonly distinguished in the literature (Piercy et al., 2018), and warrant further investigation.
Cognition.
The 12-item self-report Executive Function Index (Aharonovich et al., 2017; Spinella, 2005) assessed cognition by asking about performance of everyday tasks. Items are from two of the five subscales (Strategic Planning, Organization) of the Executive Function Index, a brief self-report measure for use in large non-clinical samples (Spinella, 2005). The ATT items assessed tendencies to lose track, lose interest, mix up the sequence of tasks, difficulty multi-tasking, and trouble summing up information for decision making purposes. The executive function items assessed planning for the future, considering consequences of an action, learning from mistakes, and self-monitoring. In a psychometric study using the NESARC-III dataset, Aharonovich and colleagues (2017) found and confirmed a two factor 9-item solution with good reliability and validity. The two factors were ATT and EF. The 9-item measure was used with higher values indicating better functioning.
2.4. Statistical Analysis
Weighed means and standard deviations were estimated for ATT and EF. Due to non-normal distributions, medians with 95% confidence intervals were calculated for the calories expended by types of PA stratified by substance use. Light-to-moderate PA, vigorous PA, and total PA scores were subsequently logarithm-transformed for analyses. Ordinary linear regression was used to regress EF and ATT scores on PA, past year substance use, and demographic variables (i.e., sex, age, race, education, personal income). Within the two facets of cognition, six separate regressions, one for each type of substance, examined the impact of light-to-moderate PA. Additional regressions separately tested vigorous PA and total PA across the different types of substances. Adjusted regression coefficients indicated associations between PA and EF and between PA and ATT by each type of substance use while controlling for demographic variables, with p<0.05 indicating significance. Significant regression coefficients for frequent or infrequent use indicated that the mean scale score differed significantly from the mean score for no use. Conditional main effects and the interaction between substance use frequency and PA were examined. If the interaction term were significant and negative, it would suggest that substance use weakens the association between PA and the cognitive domain assessed or that PA partly offsets the effect of substance use on cognition. All analyses were performed using survey procedures of SAS software version 9.4 (SAS Institute Inc., 2013). The stratification and cluster variables were used to account for the multi-stage complex sampling effects on variance estimates. Sampling weight was used to represent the U.S. population.
3. RESULTS
Demographic characteristics of the NESARC-III nationally representative dataset used in this analysis can be found in Table 1. Table 2 shows average weighted mean ATT and EF scores and median calories expended by PA intensity divided by substance use frequency. Median is reported for PA variables as a better measure of central tendency than mean due to positive skew resulting from a non-normal distribution. For individuals who endorsed never engaging in PA in the past year their calories expended per day were coded as zero. For reference, about 1,000 calories of moderate intensity PA per week is consistent with the minimum threshold for meeting the PA public health guidelines (i.e., 150 minutes of moderate intensity PA; American College of Sports Medicine, 2017).
Table 1.
Demographic characteristics of the sample
| N | Weighted % | Weighted SE | |
|---|---|---|---|
| Sex | |||
| Males | 15,862 | 48.09 | 0.30 |
| Females | 20,447 | 51.91 | 0.30 |
| Race/Ethnicity | |||
| White | 19,194 | 66.19 | 0.77 |
| African Americans | 7,766 | 11.79 | 0.66 |
| American Indian/Alaska Native | 511 | 1.56 | 0.12 |
| Asian/Native Hawaiian/Other Pacific Islander | 1,801 | 5.73 | 0.47 |
| Hispanic | 7,037 | 14.74 | 0.67 |
| Age Groups (years) | |||
| 18–24 | 4,496 | 13.05 | 0.29 |
| 25–44 | 13,765 | 34.36 | 0.39 |
| 45–64 | 12,242 | 35.03 | 0.32 |
| 65–74 | 3,358 | 10.43 | 0.23 |
| 75 or older | 2,448 | 7.13 | 0.21 |
| Education | |||
| <High School | 5,490 | 13.01 | 0.42 |
| High School/GED | 9,799 | 25.78 | 0.51 |
| Some College/Associate | 12,105 | 33.07 | 0.45 |
| College Graduates | 4,565 | 13.93 | 0.43 |
| Graduate School | 4,350 | 14.21 | 0.45 |
| Personal Income | |||
| <$20,000 | 16,928 | 44.33 | 0.56 |
| $20,000-$34,999 | 8,086 | 20.63 | 0.28 |
| $35,000-$69,999 | 7,768 | 22.49 | 0.31 |
| $70,000 or more | 3,527 | 12.54 | 0.39 |
Table 2.
Measures of central tendency of cognition and daily physical activity by frequency of substance use.
| Attention Mean (SE) |
Executive Function Mean (SE) |
Light-to-Moderate PA Median (95% CI) |
Vigorous PA Median (95% CI) |
Total PA Median (95% CI) |
|
|---|---|---|---|---|---|
| Binge drinking | |||||
| Frequent (n=4,064) | 15.86(0.07) | 9.86 (0.08) | 103.61 (92.22 – 114.99) | 216.49 (194.69 – 238.28) | 420.66 (386.89 – 454.43) |
| Infrequent (n=7,274) | 16.07 (0.06) | 10.64 (0.04) | 80.14 (75.69 – 84.59) | 157.24 (149.57 – 164.90) | 288.21 (273.26 – 303.17) |
| No use (n=23,762) | 16.07 (0.04) | 10.46 (0.04) | 59.24 (56.87 – 61.61) | 74.67 (69.32 – 80.02) | 180.58 (171.76 – 189.40) |
| Marijuana | |||||
| Frequent (n=1,858) | 14.95 (0.12) | 9.50 (0.10) | 109.75 (94.84 – 124.66) | 214.81 (180.56 – 249.07) | 405.68 (365.17 – 446.20) |
| Infrequent (n=1,754) | 15.37 (0.11) | 9.98 (0.09) | 79.78 (71.12 – 88.45) | 179.69 (160.30–199.07) | 329.88 (301.60 – 358.16) |
| No use (n=31,573) | 16.14 (0.04) | 10.51 (0.04) | 65.37 (63.08 – 67.67) | 96.58 (91.33 – 101.82) | 210.38 (202.33 – 218.44) |
| Cocaine | |||||
| Frequent (n=79) | 13.63 (0.29) | 7.31 (0.25) | 160.91 (61.72 – 260.10) | 270.61 (192.46 – 348.76) | 428.62 (315.79 – 541.45) |
| Infrequent (n=331) | 14.77 (0.24) | 8.72 (0.26) | 93.37 (65.12 – 121.63) | 174.50 (125.74 – 223.25) | 383.53 (333.59 – 433.46) |
| No use (n=34,721) | 16.06 (0.03) | 10.46 (0.04) | 67.31 (64.84 – 69.79) | 104.93 (100.01 – 109.84) | 220.65 (213.40 – 227.89) |
| Opioid | |||||
| Frequent (n=645) | 14.82 (0.24) | 8.93 (0.16) | 54.93 (42.54 – 67.33) | 68.52 (41.90 – 95.14) | 174.63 (126.09 – 223.16) |
| Infrequent (n=882) | 15.19 (0.13) | 9.67 (0.12) | 85.33 (70.96 – 99.70) | 141.23 (121.77 – 160.70) | 307.30 (262.25 – 352.36) |
| No use (n=33,593) | 16.09 (0.03) | 10.48 (0.04) | 67.34 (64.83 – 69.84) | 105.12 (100.48 – 109.75) | 220.67 (213.71 – 227.62) |
| Sedatives | |||||
| Frequent (n=345) | 14.05 (0.29) | 8.91 (0.20) | 61.62 (50.49 – 72.75) | 68.18 (51.63 – 84.73) | 177.26 (133.64 – 220.89) |
| Infrequent (n=456) | 14.65 (0.19) | 9.45 (0.14) | 80.78 (60.47 – 101.08) | 136.99 (99.22 – 174.77) | 297.49 (233.10 – 361.89) |
| No use (n=34,327) | 16.08 (0.03) | 10.47 (0.04) | 67.32 (64.83 – 69.79) | 105.89 (101.13 – 110.64) | 221.51 (214.37 – 228.65) |
| Stimulants | |||||
| Frequent (n=144) | 13.39 (0.32) | 8.89 (0.20) | 107.19 (67.29 – 147.09) | 233.66 (140.07 – 327.25) | 383.84 (254.74 – 512.94) |
| Infrequent (n=255) | 14.27 (0.23) | 9.15 (0.21) | 108.55 (69.57 – 147.53) | 283.93 (187.47 – 290.39) | 398.48 (316.69 – 480.27) |
| No use (n=34,730) | 16.07 (0.03) | 10.45 (0.04) | 67.08 (64.63 – 69.52) | 104.34 (99.41 – 109.26) | 219.95 (212.65 – 227.24) |
Note. Median is reported for physical activity due to non-normal distribution (i.e., positive skew); PA = physical activity; CI = confidence interval.
3.1. Executive Function
Ordinary linear regression was used to estimate PA and substance use effects on EF scores, controlling for demographic variables. Interaction terms between PA and substance use were also estimated. In all models the demographic covariates of age, race, education, and income were significantly associated with EF, p<.05. In some, but not all models, sex was significantly associated with EF scores. As shown in Table 3, the results of these regressions are reported by different types of substance use and by PA in participants who (1) did light-to-moderate activities, (2) vigorous activities in the past 12 months, and (3) the combination of light-to-moderate and vigorous PA. There were significant conditional main effects for all types of substance use (negative) and for total PA variables (positive) on the measure of EF (ps<.05), with the exception of infrequent binge drinking in the light-to-moderate PA group. Specifically, all types of substance use were associated with a lower EF score than no substance use among participants who did not report any PA after accounting for other covariates. The result also means that doing more PA was associated with a higher EF score among participants who did not use a substance after accounting for other covariates. For example, in the context of binge drinking the regression coefficients were 0.08 for light-to-moderate PA(log), 0.06 for vigorous PA(log), and 0.04 for total PA(log) in Table 3. Those regression coefficients mean that on average, EF increases by 0.36 point for every 100 kcal/day expended on light-to-moderate PA, EF increases by 0.27 point of for every 100 kcal/day expended on vigorous PA, and EF increases by 0.18 point for every 100 kcal/day expended on the total PA. A 100 kcal/day is equivalent to walking 1 mile for most adults.
Table 3.
Linear regression predicting executive function by daily physical activity and frequent past-year substance use.
| Binge Drinking | Marijuana Use | Cocaine Use | Opioid Use | Stimulants Use | Sedatives Use | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β | SE | P | β | SE | p | β | SE | p | β | SE | p | β | SE | p | β | SE | p | |
| Light/Moderate PA (log) | 0.08 | 0.01 | <0.001 | 0.07 | 0.03 | <0.001 | 0.07 | 0.01 | <0.001 | 0.07 | 0.01 | <0.001 | 0.07 | 0.01 | <0.001 | 0.07 | 0.01 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.46 | 0.11 | <0.001 | −0.73 | 0.16 | <0.001 | −2.26 | 0.42 | <0.001 | −0.96 | 0.20 | <0.001 | −1.77 | 0.54 | 0.001 | −1.16 | 0.24 | <0.001 |
| Infrequent use | −0.14 | 0.09 | 0.13 | −0.66 | 0.16 | <0.001 | −1.80 | 0.43 | <0.001 | −0.74 | 0.19 | <0.001 | −1.39 | 0.51 | 0.008 | −1.11 | 0.34 | .001 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | 0.07 | 0.02 | <0.001 | 0.06 | 0.03 | 0.027 | 0.15 | 0.08 | 0.042 | 0.08 | 0.04 | 0.035 | 0.06 | 0.11 | 0.60 | −0.01 | 0.04 | 0.78 |
| Infrequent use by PA | 0.01 | 0.02 | 0.57 | 0.01 | 0.03 | 0.71 | 0.00 | 0.08 | 0.99 | 0.02 | 0.04 | 0.65 | 0.00 | 0.9 | 0.99 | 0.00 | 0.07 | 0.97 |
| Vigorous PA (log) | 0.06 | 0.01 | <0.001 | 0.05 | 0.01 | <0.001 | 0.05 | 0.01 | <0.001 | 0.05 | 0.01 | <0.001 | 0.05 | .01 | <0.001 | 0.05 | 0.01 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.70 | 0.10 | <0.001 | −0.93 | 0.13 | <0.001 | −2.16 | 0.32 | <0.001 | −1.15 | 0.18 | <0.001 | −1.47 | 0.40 | <0.001 | −1.16 | 0.22 | <0.001 |
| Infrequent use | --0.18 | 0.07 | 0.018 | −0.72 | 0.14 | <0.001 | −2.01 | 0.37 | <0.001 | −0.69 | 0.20 | <0.001 | −1.38 | 0.65 | 0.036 | −0.85 | 0.26 | 0.002 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | 0.01 | 0.02 | 0.47 | 0.01 | 0.02 | 0.63 | 0.16 | 0.063 | 0.007 | 0.01 | 0.02 | 0.74 | 0.02 | 0.07 | 0.83 | −0.03 | 0.04 | 0.52 |
| Infrequent use by PA | 0.00 | 0.01 | 0.93 | 0.00 | 0.02 | 0.84 | 0.05 | 0.06 | 0.35 | 0.03 | 0.03 | 0.29 | 0.00 | 0.12 | 0.99 | −0.06 | 0.05 | 0.21 |
| Total PA (log) | 0.04 | 0.01 | <0.001 | 0.04 | 0.00 | <0.001 | 0.04 | 0.00 | <0.001 | 0.04 | 0.00 | <0.001 | 0.04 | 0.00 | <0.001 | 0.01 | 0.00 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.59 | 0.11 | <0.001 | −0.82 | 0.15 | <0.001 | −1.99 | 0.35 | <0.001 | −1.06 | 0.19 | <0.001 | −1.69 | 0.50 | 0.001 | −1.14 | 0.24 | <0.001 |
| Infrequent use | −0.18 | 0.09 | 0.036 | −0.74 | 0.17 | <0.001 | −2.00 | 0.43 | <0.001 | −0.69 | 0.22 | 0.002 | −1.36 | 0.64 | 0.037 | −0.87 | 0.33 | 0.010 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | −0.02 | 0.01 | 0.036 | −0.02 | 0.01 | 0.13 | −0.11 | 0.04 | 0.003 | −0.02 | 0.02 | 0.20 | 0.01 | 0.05 | 0.80 | −0.04 | 0.02 | 0.58 |
| Infrequent use by PA | −0.00 | 0.01 | 0.91 | 0.00 | 0.02 | 0.84 | 0.03 | 0.04 | 0.49 | −0.02 | .02 | 0.40 | −0.00 | 0.06 | 0.96 | −0.03 | 0.03 | 0.41 |
Note. PA = physical activity; Ref = reference category.
In the light-to-moderate PA models, PA was uniformly associated with significant increases on the EF score (βs range from 0.07 to 0.08) for every log-kcal increase, ps<.001. In contrast, all types of substance use were significantly associated with lower EF score when light-to-moderate PA was zero (βs range from −2.26 to −0.11). For example, the associations between the EF score and frequent binge drinking (β= −0.46, p<.001), as well as EF score and frequent marijuana use (β= −0.73, p<.001) were negative.
In the light-to-moderate PA ordinary linear regressions, significant negative interaction terms were found for frequent binge drinking, marijuana, cocaine, and opioid use suggesting that the magnitude of these positive effects for light-to-moderate PA on the EF score were significantly lower, or offset, in the context of frequent use of these substances compared to no substance use (ps<.05). For example, the interaction of light-to-moderate PA and frequent binge drinking (β= −0.07, p<.001) was associated with greater declines in EF score after accounting for other covariates. However, this multiplicative significant negative effect was not observed in the frequent stimulant and frequent sedative use regressions; nor, were any significant interactions detected within PA by infrequent use terms, ps>.05. The offset effect was larger in the frequent use than the infrequent use of substances.
After accounting for demographic and substance use, vigorous PA was uniformly associated with significant increases on the EF score (βs range from 0.05 to 0.06) for every log-kcal increase, ps<.001. Conversely, frequent substance use was uniformly associated with significant decreases in EF (βs range from −2.16 to −0.70), with frequent cocaine use associated with the most detrimental effects, ps<.001. Infrequent substance use of any substance was also associated with significantly lower EF scores (βs range from −2.01 to −0.18), with infrequent cocaine use associated with the most detrimental effects, ps<.05. The only significant interaction term detected was between vigorous PA and frequent cocaine use (β= −0.16, p=.007), which again suggests that the positive impact of vigorous PA on the EF score was significantly lower, or offset, in the context of frequent use of cocaine. Infrequent use of cocaine did not affect the association between vigorous PA and EF score.
The overall expenditure PA, meaning the combination of calories expend in light-to-moderate PA and vigorous PA, was significantly associated with higher EF scores (βs range from 0.01 to 0.04) after accounting for demographic variables, ps< .001. Consistent with models, infrequent and frequent use of any substance was significantly associated with lower EF scores (βs range from −2.0 to −0.18), ps<.05. Two interaction terms were significant in the models, ps<.05. The frequent binge drinking by total PA interaction term was negative (β=−0.02, p=.036) and so was the frequent cocaine use by total PA interaction term (β=−0.11, p=.003). This suggests that the positive effect of overall expenditure PA on the EF score was lowered, or offset, by frequent use of alcohol and cocaine. No other significant interaction between substance use and total PA was detected, ps>.05.
3.2. Attention
Table 4 shows ordinary linear regressions of ATT score predicted by PA, frequency of substance use, and the interaction between these two variables after accounting for demographic variables in participants who did light-to-moderate PA, vigorous PA in the past 12 months, and total PA. In all models the demographic covariates of sex, age, race, education, and income were significantly associated with ATT, ps<.05. Across all categories of substance use, light-to-moderate PA, vigorous PA, and the combination of light-to-moderate with vigorous PA were positively associated with ATT scores, ps<.001. All types of substance use in the past year were associated with decreases in the ATT score (βs ranged from −2.66 to −0.29), ps<.01, with the exception of infrequent use of cocaine (p=.08) and stimulants (p=.09). For example, both infrequent binge drinking and frequent binge drinking were associated with a lower ATT score (βs of −0.29 and −0.43, respectively), ps<.01.
Table 4.
Linear regression predicting attention by physical activity and frequent past-year substance use.
| Binge Drinking | Marijuana Use | Cocaine Use | Opioid Use | Stimulant Use | Sedative Use | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β | SE | p | β | SE | p | β | SE | p | β | SE | p | β | SE | p | β | SE | p | |
| Light/Moderate PA (log) |
0.04 | 0.01 | <0.001 | 0.04 | 0.01 | <0.001 | 0.04 | 0.01 | <0.001 | 0.04 | 0.04 | <0.001 | 0.04 | 0.01 | <0.001 | 0.04 | 0.01 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.34 | 0.12 | 0.005 | −1.10 | 0.16 | <0.001 | −1.79 | 0.49 | <0.001 | −0.93 | 0.27 | <0.001 | −2.75 | 0.64 | <0.001 | −1.58 | 0.32 | <0.001 |
| Infrequent use | −0.20 | 0.07 | 0.005 | −0.91 | 0.19 | <0.001 | −0.79 | 0.36 | 0.031 | −0.72 | 0.20 | <0.001 | −0.50 | 0.28 | 0.07 | −1.02 | 0.40 | 0.011 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | −0.01 | 0.02 | 0.78 | 0.00 | 0.03 | 0.92 | −0.13 | 0.08 | 0.12 | −0.03 | 0.03 | 0.30 | 0.02 | 0.14 | 0.88 | −0.11 | 0.05 | 0.06 |
| Infrequent use by PA | −0.02 | 0.02 | 0.19 | 0.03 | 0.03 | 0.40 | −0.11 | 0.06 | 0.10 | −0.04 | 0.04 | 0.33 | −0.26 | 0.07 | <0.001 | −0.09 | 0.08 | 0.25 |
| Vigorous PA (log) | 0.03 | 0.01 | <0.001 | 0.03 | 0.01 | <0.001 | 0.03 | 0.01 | <0.001 | 0.03 | 0.01 | <0.001 | 0.03 | 0.01 | <0.001 | 0.03 | 0.01 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.45 | 0.12 | <0.001 | −1.19 | 0.16 | <0.001 | −2.00 | 0.45 | <0.001 | −0.99 | 0.27 | <0.001 | −2.79 | 0.39 | <0.001 | −1.78 | 0.34 | <0.001 |
| Infrequent use | −0.31 | 0.08 | <0.001 | −1.16 | 0.18 | <0.001 | −0.99 | 0.43 | 0.022 | −0.70 | 0.21 | 0.001 | −1.25 | 0.53 | 0.020 | −1.39 | 0.32 | <0.001 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | 0.02 | 0.02 | 0.36 | 0.03 | 0.02 | 0.27 | −0.07 | 0.08 | 0.41 | 0.00 | 0.03 | 0.97 | 0.03 | 0.07 | 0.63 | −0.02 | 0.06 | 0.71 |
| Infrequent use by PA | 0.01 | 0.02 | 0.65 | 0.09 | 0.03 | 0.003 | −0.06 | 0.09 | 0.41 | −0.04 | 0.04 | 0.33 | −0.07 | 0.10 | 0.46 | 0.00 | 0.06 | 0.99 |
| Total PA (log) | 0.02 | 0.00 | <0.001 | 0.02 | 0.00 | <0.001 | 0.02 | 0.00 | <0.001 | 0.02 | 0.00 | <0.001 | 0.02 | 0.00 | <0.001 | 0.02 | 0.00 | <0.001 |
| Substance | ||||||||||||||||||
| Frequent use | −0.43 | 0.13 | 0.002 | −1.21 | 0.17 | <0.001 | −1.75 | 0.46 | <0.001 | −0.95 | 0.28 | 0.001 | −2.66 | 0.45 | <0.001 | −1.68 | 0.34 | <0.001 |
| Infrequent use | −0.29 | 0.08 | <0.001 | −1.21 | 0.20 | <0.001 | −0.80 | 0.45 | 0.08 | −0.65 | 0.24 | 0.007 | −0.65 | 0.38 | 0.09 | −1.23 | 0.42 | 0.003 |
| No use | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref. | -- | -- | Ref | -- | -- |
| Substance by PA (log) | ||||||||||||||||||
| Frequent use by PA | 0.01 | 0.01 | 0.60 | 0.01 | 0.02 | 0.40 | −0.07 | 0.05 | 0.18 | −0.01 | 0.02 | 0.66 | 0.01 | 0.05 | 0.92 | −0.04 | 0.03 | 0.31 |
| Infrequent use by PA | −0.00 | 0.01 | 0.98 | 0.05 | 0.02 | 0.006 | −0.06 | 0.04 | 0.21 | −0.03 | 0.03 | 0.28 | −0.10 | 0.04 | 0.011 | −0.02 | 0.04 | 0.65 |
Note. PA = physical activity; Ref = reference category.
Interactions between PA expenditure and substance use were also examined. Surprisingly, the interaction term of vigorous PA and infrequent marijuana use was positive and significant (β=0.09, p=.003), suggesting the positive effect of vigorous PA on ATT was significantly enhanced in the context of infrequent marijuana use. This same pattern of enhancement held for the total PA expenditure by infrequent marijuana use (β=0.05, p=.006). Meanwhile, the interaction term of infrequent stimulant use and light-to-moderate PA (β=−0.26, p<.001) was significant as was the interaction term for infrequent stimulant use and total PA (β=−0.10, p=.001). In both cases, results suggest that the positive association of PA with ATT was significantly lower, or offset, in the context of infrequent stimulant use. The offset effect of the PA expenditure on the log-transformed ATT score was greater in the frequent use of substance group than the infrequent use of substance group. No other interaction terms were significant, ps>.05.
4. DISCUSSION
This study examined the impact of and interaction between PA and substance use on cognition in a large nationally representative sample. Substance use was categorized into three frequency groups: no use in the past year, infrequent use (less than weekly), and frequent use (weekly or more often). Meanwhile, three separate variables represented PA: light-to-moderate intensity, vigorous intensity, and total PA. Analyses found greater PA engagement, no matter the intensity, was positively associated with higher self-reported cognition. To quantify these effects, a PA expenditure equivalent to daily walking of about one mile (~100 kcals/day) significantly increased EF and to a lesser degree ATT. Conversely, individuals who engaged in more frequent substance use, no matter which substance used, reported significantly poorer self-reported cognition. Even infrequent use of some substances was associated with lower self-reported cognition scores, although results were not as robust as frequent use. Thus, an individual’s engagement in health behavior has a direct positive or negative relationship to one’s cognitive abilities.
With few exceptions, we replicated the negative relationship between substance use and self-reported cognition previously reported by Aharonovich and colleagues (2017) even after incorporating an additional robust predictor (i.e., PA) of cognition into the models. Inclusion of PA is significant because the unaccounted protective effects of PA in the frequent substance use groups may have led to an underestimation of substance use’s impact. The only result that was not consistent with Aharonovich et al. (2017) was infrequent stimulant use was not associated with decreased ATT scores in two of the three regression models. More broadly, these findings are consistent with the larger literature on the negative relationship between substance use and cognition. While this study was cross-sectional, which prevents causal attribution, longitudinal studies find bi-directional relationships between substance use and cognition (Jovanovski et al., 2005; Nigg et al., 2004; Stavro et al., 2013).
Finally, the positive effect of PA expenditure on cognition was reduced in the context of substance use. Since we are unable to determine causality, results could be explained in the opposite direction as well in which PA blunts or partly offsets the negative effects of substance use on cognition. In terms of EF, four significant interaction terms were detected in the light-to-moderate PA analyses and two were found in the vigorous PA analyses. Thus, it appears individuals who engage in light-to-moderate PA are more susceptible to the damaging effects of frequent substance use on EF compared to individuals who engage in vigorous PA and frequently use substances. Alternatively, one could conclude the benefit of vigorous PA on EF is less affected by substance use than that of the light-to-moderate PA in the context of frequent binge drinking, marijuana and opioid use. The specific mechanisms of action regarding these results is unclear and warrants further investigation. Costa and colleagues (2019) provide a review of underlying neurobiological systems that may be implicated.
Overall implications of these findings are multifaceted. Clearly, PA has significant positive effects on brain health and functioning. Incorporating messages about these effects into PA interventions and public health recommendations are paramount and may be another source of motivation to facilitate initiation and maintenance of regular PA. Current messaging primarily focuses on the physical health benefits, with limited ATT to mental health benefits (Piercy et al., 2018).
Extending findings to SUD treatment, cognitive remediation and other neurocognitive approaches have recently gained momentum (Verdejo-Garcia et al., 2019). Cognition has been shown to predict poor SUD treatment adherence, treatment drop out, and relapse (Domínguez-Salas et al., 2016). Cognitive remediation interventions infrequently incorporate PA. The idea of PA as an intervention for SUD has been around for decades (Murphy et al., 1986); recently, PA was conceptualized as a non-stigmatizing intervention (Weinstock et al., 2020) that lowers the bar for treatment entry. Further evidence of the benefit of PA in the context of reducing substance use is that individuals who engage in natural recovery frequently engage in other concurrent health behavior change (i.e., begin exercising) when they reduce or stop drinking (King and Tucker, 2000). Interventions that simultaneously target PA and substance use may have a greater impact than interventions that target each behavior in isolation. A recent meta-analysis on health behavior change interventions found targeting two to three health behaviors (e.g., PA, substance use) had larger effect sizes than those targeting one or greater than three health behaviors at a time (Wilson et al., 2015). Therefore, framing a PA intervention that also seeks to reduce substance use as a “wellness” intervention may have beneficial impacts on substance use, cognition, and other aspects of physical and mental health (e.g., reductions in blood pressure and depression). We caution against initially prescribing vigorous PA as part a cognitive remediation intervention. Vigorous PA may result in poor adherence due to limited executive control both before and during PA potentially leading to a low threshold for fatigue and tolerance of discomfort (Costa et al., 2019).
4.1. Limitations
This study is not without limitations. Primary among them is that PA was assessed via self-report. Objective measures (e.g., accelerometer, cardiorespiratory fitness) are commonly used in examining the PA-cognition relationship and provide more precise measurement of PA/fitness (Stillman et al., 2020). Moreover, objective measures are not subject to under or over-reporting and recall bias (Panza et al., 2012). Additionally, the PA questionnaire included in the NESARC battery combined the assessment of light intensity PA with moderate intensity PA into one item. This combined item resulted in a conservative estimation of calories expended per day to account for the possibility that the PA reported was light (and not moderate). Moreover, the combination of light and moderate intensities precludes investigation of PA that is consistent with public health guidelines (e.g., ≥150 minutes per week of moderate intensity PA) or examination of any differential effects between light and moderate PA. Self-reports of cognition may also be perceived as a limitation of this study; yet, self-reports of cognition have demonstrated incremental validity over performance-based measures of cognition in predicting outcomes such as functional impairment (Hagen et al., 2016). Prior work has validated the measure used in this study (Aharonovich et al., 2017; Spinella, 2005) and brief self-reports are the preferred method in large epidemiological studies. Correction for multiple comparisons were not utilized and may be considered a limitation. Type I error (false positives) may have occurred and replication in additional samples is warranted. Finally, data are cross-sectional and causal inferences cannot be determined.
4.2. Conclusion
In summary, greater PA was associated with higher self-reported cognition, and more frequent substance use was related to lower self-reported cognition. When the interactions between PA and substance use were examined, the positive impact of PA on EF was reduced in the context of substance use across most substances. Findings suggest that individuals that participate in light-to-moderate PA and engage in more frequent substance use may experience worse cognitive functioning than those who engage in vigorous PA and frequently use substances. Thus, findings support the notion that cognition can be influenced by health behaviors, and aerobic activity may serve as a protective behavior that preserves cognition. These findings have implications for improving SUD treatment as well as adjusting public health recommendations for PA to include benefits to cognition.
Highlights.
The impact of moderate to vigorous PA on cognition is inconclusive in adults.
We found consistent positive association between PA and self-reported cognition.
Frequent substance use was detrimental to cognition.
Frequent substance use lessened the PA’s positive effects on cognition.
Vigorous PA’s cognition effects appears to be less susceptible to substance use
Role of Funding Source:
Research reported in this publication was supported by the National Institute on Drug Abuse of the National Institutes of Health under Award Number R01DA033411. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This research was supported by the Department of Veterans Affairs Office of Academic Affiliations Advanced Fellowship Program in Mental Illness Research and Treatment and the San Francisco VA Health Care System.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest: No conflict declared.
References
- Aharonovich E, Shmulewitz D, Wall MM, Grant BF, Hasin DS, 2017. Self-reported cognitive scales in a US National Survey: reliability, validity, and preliminary evidence for associations with alcohol and drug use. Addiction 112(12), 2132–2143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Allegri N, Mennuni S, Rulli E, Vanacore N, Corli O, Floriani I, De Simone I, Allegri M, Govoni S, Vecchi T, Sandrini G, Liccione D, Biagioli E, 2019. Systematic review and meta-analysis on neuropsychological effects of long-term use of opioids in patients with chronic noncancer pain. Pain Pract 19(3), 328–343. [DOI] [PubMed] [Google Scholar]
- American College of Sports Medicine, 2017. ACSM’s exercise testing and prescription, 10th ed. Lippincott Williams & Wilkins. [Google Scholar]
- Audiffren M, André N, 2019. The exercise–cognition relationship: A virtuous circle. J Sport Health Sci 8(4), 339–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Costa KG, Cabral DA, Hohl R, Fontes EB, 2019. Rewiring the addicted brain through a psychobiological model of physical exercise. Front Psychiatry 10(600). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Diamond A, 2013. Executive functions. Annu Rev Psychol 64(1), 135–168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Domínguez-Salas S, Díaz-Batanero C, Lozano-Rojas OM, Verdejo-García A, 2016. Impact of general cognition and executive function deficits on addiction treatment outcomes: Systematic review and discussion of neurocognitive pathways. Neurosci Biobehav Rev 71, 772–801. [DOI] [PubMed] [Google Scholar]
- Erickson KI, Hillman C, Stillman CM, Ballard RM, Bloodgood B, Conroy DE, Macko R, Marquez DX, Petruzzello SJ, Powell KE, For Physical Activity Guidelines Advisory, C., 2019. Physical activity, cognition, and brain outcomes: A review of the 2018 Physical Activity Guidelines. Med Sci Sports Exer 51(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Galaj E, Barrera ED, Ranaldi R, 2020. Therapeutic efficacy of environmental enrichment for substance use disorders. Pharmacol Biochem Behav 188, 172829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grant BF, Goldstein RB, Saha TD, Chou SP, Jung J, Zhang H, Pickering RP, Ruan WJ, Smith SM, Huang B, 2015a. Epidemiology of DSM-5 alcohol use disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions III. JAMA Psychiatry 72(8), 757–766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grant BF, Goldstein RB, Smith SM, Jung J, Zhang H, Chou SP, Pickering RP, Ruan WJ, Huang B, Saha TD, Aivadyan C, Greenstein E, Hasin DS, 2015b. The Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (AUDADIS-5): Reliability of substance use and psychiatric disorder modules in a general population sample. Drug Alcohol Depend 148, 27–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hagen E, Erga AH, Hagen KP, Nesvåg SM, McKay JR, Lundervold AJ, Walderhaug E, 2016. Assessment of executive function in patients with substance use disorder: A comparison of inventory- and performance-based assessment. J Subst Abuse Treat 66, 1–8. [DOI] [PubMed] [Google Scholar]
- Head D, Singh T, Bugg JM, 2012. The moderating role of exercise on stress-related effects on the hippocampus and memory in later adulthood. Neuropsychology 26(2), 133–143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jovanovski D, Erb S, Zakzanis K, 2005. Neurocognitive deficits in cocaine users: A quantitative review of the evidence. J Clin Exp Neuropsychol 27(2), 189–204. [DOI] [PubMed] [Google Scholar]
- Kara KP, Matthew WM, Leah ER, 2013. Acute exercise enhances preschoolers’ ability to sustain attention. J Sport Exerc Psychol 35(4), 433–437. [DOI] [PubMed] [Google Scholar]
- King MP, Tucker JA, 2000. Behavior change patterns and strategies distinguishing moderation drinking and abstinence during the natural resolution of alcohol problems without treatment. Psychol Addict Behav 14, 48–55. [DOI] [PubMed] [Google Scholar]
- Krell-Roesch J, Syrjanen JA, Vassilaki M, Barisch-Fritz B, Trautwein S, Boes K, Woll A, Kremers WK, Machulda MM, Mielke MM, Knopman DS, Petersen RC, Geda YE, 2019. Association of non-exercise physical activity in mid- and late-life with cognitive trajectories and the impact of APOE ε4 genotype status: the Mayo Clinic Study of Aging. Eur J Ageing 16(4), 491–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Le Berre A-P, Fama R, Sullivan EV, 2017. Executive functions, memory, and social cognitive deficits and recovery in chronic alcoholism: A critical review to inform future research. Alcohol Clin Exp Res 41(8), 1432–1443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee PH, Macfarlane DJ, Lam TH, Stewart SM, 2011. Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. Int J Behav Nutr Phys Act 8(1), 115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Manning V, Verdejo-Garcia A, Lubman DI, 2017. Neurocognitive impairment in addiction and opportunities for intervention. Curr Opin Behav Sci 13, 40–45. [Google Scholar]
- Masley S, Roetzheim R, Gualtieri T, 2009. Aerobic exercise enhances cognitive flexibility. J Clin Psychol Med Settings 16(2), 186–193. [DOI] [PubMed] [Google Scholar]
- Moreau D, Chou E, 2019. The acute effect of high-intensity exercise on executive function: A meta-analysis. Perspect Psychol Sci 14(5), 734–764. [DOI] [PubMed] [Google Scholar]
- Morin J-FG, Afzali MH, Bourque J, Stewart SH, Séguin JR, O’Leary-Barrett M, Conrod PJ, 2018. A population-based analysis of the relationship between substance use and adolescent cognitive development. Am J Psychiatry 176(2), 98–106. [DOI] [PubMed] [Google Scholar]
- Mulhauser K, Weinstock J, Ruppert P, Benware J, 2018. Changes in neuropsychological status during the initial phase of abstinence in alcohol use disorder: Neurocognitive impairment and implications for clinical care. Subst Use Misuse 53(6), 881–890. [DOI] [PubMed] [Google Scholar]
- Murphy TJ, Pagano RR, Marlatt GA, 1986. Lifestyle modification with heavy alcohol drinkers: Effects of aerobic exercise and meditation. Addict Behav 11, 175–186. [DOI] [PubMed] [Google Scholar]
- Nigg JT, Glass JM, Wong MM, Poon E, Jester JM, Fitzgerald HE, Puttler LI, Adams KM, Zucker RA, 2004. Neuropsychological executive functioning in children at elevated risk for alcoholism: Findings in early adolescence. J Abnorm Psychol 113(2), 302–314. [DOI] [PubMed] [Google Scholar]
- Opel N, Martin S, Meinert S, Redlich R, Enneking V, Richter M, Goltermann J, Johnen A, Dannlowski U, Repple J, 2019. White matter microstructure mediates the association between physical fitness and cognition in healthy, young adults. Sci Rep 9(1), 12885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Panza GA, Weinstock J, Ash GI, Pescatello LS, 2012. Psychometric evaluation of the Timeline Followback for Exercise among college students. Psychol Sport Exerc 13(6), 779–788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Piercy KL, Troiano RP, Ballard RM, Carlson SA, Fulton JE, Galuska DA, George SM, Olson RD, 2018. The physical activity guidelines for Americans. JAMA 320(19), 2020–2028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ramey T, Regier PS, 2019. Cognitive impairment in substance use disorders. CNS Spectr 24(1), 102–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sampedro-Piquero P, Ladrón de Guevara-Miranda D, Pavón FJ, Serrano A, Suárez J, Rodríguez de Fonseca F, Santín LJ, Castilla-Ortega E, 2019. Neuroplastic and cognitive impairment in substance use disorders: a therapeutic potential of cognitive stimulation. Neurosci Biobehav Rev 106, 23–48. [DOI] [PubMed] [Google Scholar]
- SAS Institute Inc., 2013. Base SAS 9.4 Procedures Guide: Statistical Procedures Cary, NC, USA: SAS Institute Inc. [Google Scholar]
- Spinella M, 2005. Self-rated executive function: Development of the Executive Function Index. Int J Neurosci 115(5), 649–667. [DOI] [PubMed] [Google Scholar]
- Stavro K, Pelletier J, Potvin S, 2013. Widespread and sustained cognitive deficits in alcoholism: A meta-analysis. Addict Biol 18(2), 203–213. [DOI] [PubMed] [Google Scholar]
- Stillman CM, Esteban-Cornejo I, Brown B, Bender CM, Erickson KI, 2020. Effects of exercise on brain and cognition across age groups and health states. Trends Neurosci 43(7), 533–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Verdejo-Garcia A, Lorenzetti V, Manning V, Piercy H, Bruno R, Hester R, Pennington D, Tolomeo S, Arunogiri S, Bates ME, Bowden-Jones H, Campanella S, Daughters SB, Kouimtsidis C, Lubman DI, Meyerhoff DJ, Ralph A, Rezapour T, Tavakoli H, Zare-Bidoky M, Zilverstand A, Steele D, Moeller SJ, Paulus M, Baldacchino A, Ekhtiari H, 2019. A roadmap for integrating neuroscience into addiction treatment: A consensus of the Neuroscience Interest Group of the International Society of Addiction Medicine. Front Psychiatry 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weinstock J, Petry NM, Pescatello LS, Henderson CE, Nelson CR, 2020. Randomized clinical trial of exercise for nontreatment seeking adults with alcohol use disorder. Psychol Addict Behav 34(1), 65–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wilson K, Senay I, Durantini M, Sánchez F, Hennessy M, Spring B, Albarracín D, 2015. When it comes to lifestyle recommendations, more is sometimes less: A meta-analysis of theoretical assumptions underlying the effectiveness of interventions promoting multiple behavior domain change. Psycho Bull 141(2), 474–509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wollman SC, Hauson AO, Hall MG, Connors EJ, Allen KE, Stern MJ, Stephan RA, Kimmel CL, Sarkissians S, Barlet BD, Flora-Tostado C, 2019. Neuropsychological functioning in opioid use disorder: A research synthesis and meta-analysis. Am J Drug Alcohol Abuse 45(1), 11–25. [DOI] [PubMed] [Google Scholar]