Grit in Patients with Substance Use Disorders

Abstract

Background and Objectives

Grit is an emerging concept in positive psychology, defined as the ability to be persistent and focused in pursuit of long-term goals. This concept has received a great deal of interest recently because of its robust ability to predict success and well-being across a wide variety of domains. The study aim was to examine the clinical relevance of the construct of grit among patients with substance use disorders.

Methods

Inpatients on a detoxification unit were enrolled from September 2013-August 2015 (N=673). Psychometric properties of the Short Grit Scale (Grit-S) were reported. We then examined sociodemographic and clinical variables that might be associated with grit in this population.

Results

In this sample of patients with substance use disorders, the total Grit-S demonstrated strong psychometric properties. Grit-S scores were higher among older patients and those who were employed; scores were lower among those never married, diagnosed with a co-occurring psychiatric disorder, or who had used heroin during the past month, according to bivariate analyses. Grit-S scores remained associated with age, employment, and presence of a co-occurring psychiatric disorder in adjusted analysis.

Conclusions and Scientific Significance

This study provides initial support for the utility of the Grit-S among those with substance use disorders; this novel measure has not been previously reported in clinical populations. Research examining grit prospectively is needed to determine whether the links between grit and outcomes observed in other populations apply to patients with substance use disorders.

“Nothing in the world can take the place of persistence. Talent will not; nothing is more common than unsuccessful men with talent. Genius will not; unrewarded genius is almost a proverb. Education will not; the world is full of educated derelicts. Persistence and determination are omnipotent.”

(Attributed to Calvin Coolidge)

INTRODUCTION

Substance use disorders are chronic conditions that require sustained and persistent effort to achieve long-term recovery. Recovery from a substance use disorder requires perseverance in dealing with a multitude of obstacles associated with relapse, such as acute and protracted withdrawal¹, social network influences², environmental triggers of craving³, and stress⁴.

The trait-level ability to be persistent and focused in the pursuit of long-term goals⁵ is known as “grit;” this positive-psychology concept has received a great deal of interest recently because of its robust ability to predict success and well-being across a wide variety of domains. People with high levels of grit persevere despite obstacles, maintain long-term focus, and see difficult tasks through to their completion⁶. Research on grit has focused predominantly on its association with academic achievement. In a series of studies examining grit in a variety of educational settings, Duckworth and colleagues⁵ found that university students with higher levels of grit achieved higher grade point averages than their peers, adjusted for college admission test scores (Scholastic Aptitude Test, or SAT). Grit was also associated with performance in the Scripps National Spelling Bee, adjusted for conscientiousness⁶; and was a better predictor of retention among West Point students (i.e., The U.S. Military Academy at West Point) following their first summer of rigorous military training than self-control, conscientiousness, or a composite measure of high school rank, SAT score, leadership potential, and physical aptitude⁵. Educators have come to view grit as a key skill to develop for the rigors of academic and professional goal-striving⁷; indeed, early efforts to increase students’ ability to persevere in goal pursuit have shown promise in improving academic performance⁸. Grit appears to capture a specific ability to persist through particularly difficult situations.

The importance of grit extends beyond the academic setting. In studies examining non-student populations, those with lower grit had less employment and relationship stability, including greater job attrition⁹, more frequent career changes⁶, and, among men, greater likelihood of divorce⁹. Furthermore, several studies have linked higher grit to greater psychological well-being, including lower suicidal ideation¹⁰ and lower rates of burnout during surgical residency¹¹. Those with higher levels of grit may have a healthier outlook on life: grit has been shown to be associated with work engagement¹² and is associated with an orientation toward pursuing engagement and meaning in life¹³. Thus, grit appears to be linked to a wide array of outcomes reflecting both achievement and well-being. The links between grit and psychological health (e.g., suicidal ideation) highlight this construct as potentially relevant to psychiatric disorders.

No published studies to date have examined grit in a clinical sample; grit may be relevant not only for achievement and well-being, but also for perseverance and sustained interest during the process of recovery. The ability to sustain interest in the goal of recovery over the long-term is especially pertinent to those with a chronic problem such as a substance use disorder. Patients with substance use disorders need to abstain from substances of abuse day after day, despite challenges presented by stresses, psychological and physical pain, and cravings. The ability to sustain an abstinence goal and to persist in its achievement appears to thus be quite relevant for this patient population.

Patients with substance use disorders have exhibited deficits in areas that may overlap with grit, including self-control, studied extensively in patients with substance use disorders^14,15, and conscientiousness¹⁶. Although grit is associated with both traits, grit has been shown to have predictive validity for successful outcomes beyond these concepts^5,6 and may be particularly important in this population due to its relevance in the pursuit of arduous goals. Grit, however, has not been examined in a population with substance use disorders; it is therefore necessary to determine if methods of measuring grit are psychometrically sound in this population, and if grit is associated with similar characteristics as it is in non-clinical samples. It is further necessary to examine whether grit is distinct from other constructs that characterize patients with substance use disorders, such as craving and impulsivity, which may also contribute to difficulties with goal pursuit.

The overarching aim of this exploratory study was to evaluate the clinical relevance of grit in a sample of patients with substance use disorders. First, we examined the psychometric properties of the Short Grit Scale (Grit-S), a self-report measure of grit that has been validated in non-clinical and unselected samples⁶; our aim was to determine whether the Grit-S was reliable and valid among treatment-seeking patients with substance use disorders. Given a recent meta-analysis of the Grit-S, largely in student populations, that found subscale differences¹⁷, we examined the psychometric properties of each subscale. We then examined whether grit was associated with markers of clinical severity and functioning among patients with substance use disorders. We hypothesized that low grit scores would be correlated with markers of difficulties in recovery, such as poor employment history and co-occurring psychiatric illness.

METHODS

Participants

This study was conducted in a 17-bed inpatient detoxification unit, which admits approximately 25 patients per week for an average of 4 days. Patients were invited to participate in a self-administered survey, completed on paper initially (n=189) and later via tablet computer (n=513). This study was approved by the McLean Hospital Institutional Review Board. Written informed consent was obtained after a complete study description. Of those approached September 2013-August 2015, 67.9% (704/1034) were enrolled; however, one of these was determined to be ineligible (having already completed the survey during a prior admission) and another one provided inconsistent data. Those who did not participate either refused (n=152), postponed participation and were then discharged (n=175), were ineligible (n=2), or withdrew consent (1). Reasons for refusal were not interested (n=97), privacy concerns (n=24), imminent discharge (n=25), and feeling unwell (n=6). The sample size for the current analysis was N=673, as 29 patients failed to complete the Grit-S.

Measures

The Short Grit Scale (Grit-S)⁶, an 8-item measure of grit, provides a total score and two subscale scores: Perseverance of Effort (sample item: “Setbacks don’t discourage me”) and Consistency of Interest (sample item, reverse scored: “New ideas and projects sometimes distract me from previous ones”). Scores range from 1 to 5, with higher scores indicating more grit. The Grit-S has demonstrated moderate internal consistency across several samples, with Cronbach’s alphas ranging from .73 to .83 for the total scale, .73 to .79 for the Consistency of Interest subscale, and .60 to .78 for the Perseverance of Effort subscale⁶. Test-retest reliability has been reported with correlations over 4–6 months of .61–.68 in student samples^6,18.

Demographic information, primary DSM-IV substance use disorder diagnosis, and additional DSM-IV psychiatric diagnoses were obtained from participants’ medical records. To establish that grit was not simply reflective of more severe substance use disorder, greater impulsivity, or stronger craving, several self-report questionnaires were collected to assess these characteristics.

The Brief Addiction Monitor¹⁹ is a 17-item measure of substance use disorder severity, including days of substance use, risk factors for poor prognosis (e.g., negative mood, craving, and arguments with friends or family), and protective factors (e.g., attending self-help meetings, spending time with supportive people, and working). Items are ordinal, ranging from 0 to 4.

Impulsivity was assessed by the 27-item, self-administered Monetary-Choice Questionnaire²⁰. This measure examines delay discounting, or the tendency to overvalue immediate relative to more distant rewards, and is a risk factor for the development of substance use disorders²¹. Each item requires the respondent to choose between smaller, immediate monetary rewards and larger, delayed monetary rewards. Scores range from 0 to 0.25, with higher scores reflecting a greater discounting of delayed rewards, hence more impulsivity. Test-retest reliability is strong, ranging between 0.77 for five weeks and 0.71 for one year²².

The Craving Scale²³ is a 3-item measure of craving for a participant’s primary drug of abuse. Items measure desire to use the drug, likelihood of using it in the environment in which previous use occurred, and cue-induced urges to use, each on a 10-point scale (0 to 9). Scale scores range from 0 to 27, with higher scores indicating greater craving. The Craving Scale has previously demonstrated good internal consistency (α=.78)²³ and predictive validity for drug use among those with opioid³ and other substance use disorders, including cocaine²³.

Statistical Analysis

Psychometric properties assessed for the Grit-S included internal consistency (using Cronbach’s alpha) and discriminant validity (using correlation coefficients) for the 673 out of 702 participants with complete Grit-S data. Pearson correlation coefficients, independent t-tests, or one-way analysis of variance (with post-hoc t-tests) compared Grit-S scores by sociodemographic and clinical characteristics, followed by linear regression to examine these associations in an adjusted model. The adjusted model included gender and those characteristics with p-values <.05 in bivariate analysis (only the characteristic of primary opioid use disorder was excluded because it was redundant with heroin use). Data were analyzed using SPSS v.20.

RESULTS

Sample Description

The 673 participants were 18 to 82 years old, with a mean age of 37.6 (sd=14.2). Nearly all (93.5%, n=629) participants were white, and two-thirds (68.5%, n=461) were male. Almost half (46.2%, n=311) were employed, and most (53.2%, n=358) were never married. Most participants were admitted for primary alcohol use disorder (58.4%, n=393), followed by opioid use disorder (39.8%, n=268); the remainder had benzodiazepine, cocaine, or sedative use disorder (n=12). The majority of participants (76.8%, n=517) used their primary substance 16–30 days in the month prior to study participation. Two-thirds (67.8%, n=456) of participants had at least one co-occurring psychiatric disorder other than a substance use disorder; primarily depressive disorders (66.9%, n=305), followed by posttraumatic stress disorder (7.0%, n=32) and other anxiety disorders (15.4%, n=70), bipolar disorder (5.9%, n=27), and other disorders (4.8%, n=22).

Psychometric Properties of the Grit-S

We assessed the Grit-S, using the SMOG Readability formula²⁴. Reading grade level was estimated to be ninth grade. The total Grit-S displayed good internal consistency overall (α=.74). Consistent with findings in other populations⁶, the Consistency of Interest subscale had excellent internal consistency (α=.81), whereas the Perseverance of Effort subscale had lower internal consistency (α=.60). Scores for the total Grit-S were approximately normally distributed (range from 1.2–4.8), with both the mean and the median=3.1 and no discernible skew; subscales were similarly distributed. We examined correlations between the Grit-S and other distinct constructs to assess discriminant validity. The Grit-S demonstrated good discriminant validity, as all correlations with measures of impulsivity (r=−0.13, n=324), craving (r=−0.25, n=658), and risk and protective factors in recovery from substance use disorder (r’s≤0.19, n=626–659) were low. Discriminant validity for the subscales was also supported (r’s<.20 for consistency and <.23 for perseverance, all in the same direction as the total Grit-S. The subscale correlation was low but significant (r=.30, p<.001). A factor analysis with oblique rotation (promax) confirmed the two-factor structure of the scale, with eigenvalues=3.16 (39.5% of the variance explained) for consistency and 1.51 (18.9%) for perseverance.

Comparison of Grit-S scores to Other Samples

Compared to a general sample of adults age ≥25 who participated in a voluntary online study of the Grit-S⁶ (N=1554), our study sample of inpatients with substance use disorder had lower Grit-S scores (means=3.1 vs. 3.4; t(672)= −10.44, p<0.001). In fact, our sample scored significantly lower (p<0.001) on the Grit-S than reports of other samples (means=3.4–3.9) of volunteers online^9,13 and high-achieving subjects^11,25, with one exception (mean=2.8 in another sample of online volunteers¹³).

Sociodemographic and Clinical Variables Associated with the Grit-S: Bivariate Analysis

We examined the association between demographic and clinical variables and grit in a bivariate analysis (Table 1). Grit-S scores were higher for older participants, those ever married, and those employed, whereas gender and race were not associated with Grit-S scores. Frequent use of the primary substance in the past month (defined as >15 days) and prior detoxification were not associated with Grit-S scores, but use of any heroin in the past month and the presence of a co-occurring psychiatric disorder were both associated with lower Grit-S scores. Frequency using the primary substance in the past month was also not associated with grit when measured on a 5-point ordinal scale. The effect sizes (Table 1) of these statistically significant associations are considered to be small to medium²⁶.

TABLE 1.

Sociodemographic and clinical characteristics associated with the Grit-S

Participant characteristic		Mean	SD	d	t/F	p-value
Sociodemographica
Gender	Male (n=461)	3.15	0.65	.07	−0.82	.41
	Female (n=209)	3.10	0.69
Race	White (n=629)	3.13	0.66	−.09	0.53	.60
	Other (n=35)	3.19	0.69
Marital status	Never married (n=358)	3.01	0.64	−.41	5.45	<.001
	Other (n=314)	3.28	0.66
Employment	Yes (n=311)	3.25	0.66	.34	−4.44	<.001
	No (n=345)	3.03	0.65
Clinical
Used primary substance, past monthb	≤15 days (n=155)	3.13	0.64	<.01	0.02	.99
	>15 days (n=517)	3.13	0.67
Used heroin, past month	Yes (n=230)	2.98	0.63	−.36	4.48	<.001
	No (n=443)	3.22	0.66
Co-occurring psychiatric disorder	Present (n=456)	3.07	0.65	−.28	3.42	.001
	Absent (n=217)	3.26	0.67
Prior detoxification	Yes (n=297)	3.16	0.65	.10	−1.10	.27
	No (n=196)	3.09	0.68
Primary substance use disorder c
	Alcohol (393)	3.21	0.68	8.62	<.001
	Opioids (268)	3.04	0.62
	Other (12)	2.66	0.53

^aFor the continuous variable of age, effect size r=.24, p<.001

^bFor frequency used primary substance analyzed as an ordinal variable, r=−.03, p=.50.

^c

Effect sizes & post hoc comparisons:

• for alcohol vs. opioid use disorder, ES=.18; t(659)=3.28, p=.001; and
• for alcohol vs. other substance use disorder, ES=.60; t(403)=2.81, p=.01

Primary substance use disorder diagnosis was also associated with Grit-S scores (Table 1): participants with primary alcohol use disorder had higher scores than those with primary opioid use disorder. Since patients whose primary substance of abuse was heroin have been shown to have worse clinical presentation than those who primarily abuse prescription opioids²⁷, we compared Grit-S scores by type of opioid use disorder. Indeed, heroin users had the lowest scores: the discrepancy in grit between those with alcohol versus opioid use disorder was larger for those who primarily used heroin (defined as use of heroin ≥4 days and prescription opioids <4 days in the month prior to treatment entry) or those who used a combination of heroin and prescription opioids (defined as use of both drugs ≥4 days in the past month). Participants with primary alcohol use disorder had higher Grit-S scores than those who primarily used heroin (n=393, mean (sd)=3.21 (0.68) vs. n=114, mean (sd)=3.00 (0.66); t(505)= −2.97, p=.003) and those who used a combination of heroin and prescription opioids (n=89, mean (sd)=2.98 (0.58); t(480)= −2.99, p=.003). There were no differences between those with alcohol use disorder and those who primarily abused prescription opioids (i.e., used heroin <4 days in the past month; n=61, mean (sd)=3.19 (0.56); t(452)= −0.20, p<.84). Because only 12 participants were diagnosed with a primary substance use disorder other than alcohol or opioid use disorder, statistical analysis of these participants would be unwarranted. Subscale analysis showed the same pattern for perseverance; however, consistency did not vary by primary substance use disorder diagnosis.

Sociodemographic and Clinical Variables Associated with the Grit-S: Multivariable Analysis

Multivariable regression was used to assess the relative contribution of sociodemographic and clinical variables associated with grit: gender plus the characteristics with p-values <.05 in bivariate analysis. Primary substance use disorder was excluded from the model because of overlap with heroin use. In the adjusted model (Table 2), age, employment status, and co-occurring psychiatric disorder were associated with Grit-S scores. Participants who were older or employed had higher Grit-S scores, whereas those with a co-occurring psychiatric disorder had lower Grit-S scores. The effect sizes (Table 2) of these statistically significant associations are considered to be small²⁶. Neither heroin use nor marital status was associated with total Grit-S scores in the adjusted model, as these bivariate associations were confounded with age (for heroin use and age, r=−0.52; for never married and age, r=−.60).

TABLE 2.

Sociodemographic and clinical predictors of Grit-S score in a linear regression model of patients with substance use disorder (N=652)

Participant characteristic	Ba	SE	da	t	p-value
Sociodemographic
Male	.05	.05	.08	0.94	<.35
Age	.01	<.01	.13	3.44	.001
Never married	−.10	.06	−.15	−1.55	.12
Employed	.17	.05	.27	3.36	.001
Clinical
Heroin use, past month	−.08	.06	−.12	−1.23	<.22
Co-occurring psychiatric disorder	−.20	.05	−.31	−3.64	<.001

^aCohen’s d for dichotomous variables and a partial correlation for age

DISCUSSION

To investigate the potential clinical relevance of grit in patients with substance use disorders, we examined the association of the Grit-S with demographic and clinical characteristics in a sample of 673 inpatients with substance use disorders. The total Grit-S demonstrated strong psychometric properties, providing initial support for its use in this population. Adjusted analyses showed that lower Grit-S scores were associated with younger age, unemployment, and presence of a co-occurring psychiatric disorder. Interestingly, among patients with substance use disorder, younger age^28,29 and co-occurring disorders²⁸ have been associated with worse substance use outcomes.

Although the present study is the first to examine grit in inpatients with substance use disorders, the findings complement previous research on related constructs in this population. Those with substance use disorders demonstrate deficits in self-control¹⁵ and conscientiousness¹⁶, characteristics associated with, yet distinct from, grit^5,6. Substance use disorders are further characterized by a number of cognitive control problems, including deficits in complex decision-making³⁰ and delaying gratification³¹. It is not surprising that inpatients with substance use disorders have lower grit than non-clinical samples^6,9,13 given results in related domains: greater impulsivity and a preference for short-term rewards³² point to a present-focused mindset, one characterized by poor future planning.

Related constructs, including impulsivity and reward processing, do not consistently predict substance use disorder treatment outcome^33,34; hence the construct of grit may be promising. Grit is unique in focusing on the ability to pursue a goal over the long-term, which may be relevant to the ability to adhere to a treatment regimen or to sustain abstinence. As demonstrated in a non-clinical population, grit can be a better predictor of successful goal pursuit than related concepts, such as self-control⁵.

Variability within our sample merits consideration. In particular, employment and co-occurring psychiatric disorder both predicted grit scores in adjusted analysis, consistent with studies associating higher grit with greater job retention⁹ and psychological well-being¹³, as well as lower suicidal ideation and less severe depressive symptoms in non-clinical populations¹⁰. Further, the magnitude of effect sizes was similar to those reported in other studies in which grit scores differentiated between groups in important outcomes, for example, between West Point cadets who dropped out during basic training and those who continued³⁵. Thus, in this clinical sample, variations in grit aligned with previous research on this construct. Further, the effect sizes of the associations between grit and clinical characteristics were of a small to medium magnitude. This magnitude of effect is not surprising for a study comparing participants within a clinical sample (i.e., inpatients receiving treatment for substance use disorders), and is comparable to the effects found for other clinically meaningful trait-level variables when comparing patients within substance use disorders samples (e.g., anxiety sensitivity)^36,37. Larger differences in grit may characterize comparisons between more heterogeneous populations (e.g., social drinkers vs. those with alcohol use disorder; adults with and without substance use disorders); future studies are needed to determine such differences.

Although scores may vary by context, the Grit-S showed good test-retest reliability in student samples^6,18. Future research could assess grit longitudinally in a clinical population to examine variability over time and effects of life events. Such research could also provide evidence as to whether grit is constant across tasks or is context-specific, as is true for certain other personality traits, such as perfectionism³⁸, distress tolerance³⁹, and impulsivity⁴⁰. For example, a person may demonstrate grit in pursuit of career goals but not in the pursuit of abstinence. Clarifying the nature of grit in various settings could be useful in understanding this emerging concept and in creating interventions to increase grit in those with broad or potentially context-specific deficits.

This study has several limitations. Since validated behavioral measures of grit are not available, our study relied exclusively on self-report assessment. Second, since participants were inpatients on a detoxification unit and many had recently relapsed, they may have interpreted some Grit-S items (e.g., “Setbacks don’t discourage me”) differently from respondents whose substance use disorder was more stable or respondents in non-clinical samples, who may not have completed the questionnaire in the context of a major setback. Finally, this study was cross-sectional in design and thus prospective associations between grit and clinical variables of interest (e.g., relapse) cannot be established. The current paper represents a first step to evaluate the reliability and validity of this measure in a clinical sample in order to justify its use in longitudinal studies. Because we did find some differences between Grit-S scores and subscale scores, and a lower correlation between the two subscales than that reported for other populations (r=.66) ¹⁷, future research in substance use disorder patients might examine the subscales as well as the total Grit-S.

Longitudinal designs examining the impact of grit over time are needed to establish the predictive validity of grit for substance use disorder outcomes. For example, studies examining whether grit is associated with treatment dropout and duration of abstinence will help to determine if grit might be a valuable target for intervention to enhance outcomes.

Considering the relevance of perseverance and focus to the recovery process, the enhancement of grit may be an important target to consider in substance use disorder treatment. Indeed, stressing the importance of grit, i.e., the importance of not giving up despite obstacles and continuing to focus on the key goals of recovery, could be a helpful adjunct to standard substance use disorder treatment. If the assertion in the meta-analysis of Credé and colleagues¹⁷ is correct, i.e., that “grit may be most useful when the task is difficult but well-defined,” then recovery from substance use disorder appears to be a good fit. In particular, those younger in age, unemployed, or with a co-occurring psychiatric disorder might benefit from a treatment that included a focus on increasing grit. Although no interventions to increase grit in a clinical sample have been published, the technique of Duckworth and colleagues⁸ to teach young students to identify plans to overcome obstacles in goal pursuit has been shown to moderately increase school attendance and academic performance. Furthermore, interventions to improve other self-regulation traits, including delay discounting⁴¹, self-control capacity⁴² and distress tolerance⁴³, have previously been efficacious. Whether grit can similarly be increased in patients with substance use disorders and whether increasing grit could improve treatment outcomes are promising areas for future research.