Introduction

Exercise is often viewed as positive and adaptive, as it is associated with various physical and mental improvements [1]. Some exercise, however, is damaging to both physical and mental health and is associated with increased risk for physical injury and the development of psychological disorders (e.g., eating disorders) [2, 3]. Definitions of maladaptive exercise highlight the interplay between the negative cognitive and behavioral aspects of exercise, suggesting that high-frequency exercise alone may not necessarily be problematic [4]. As such, it is important to explore additional factors such as exercise motivations, that may contribute to whether exercise behavior is maladaptive. Exercise motivations have been investigated in relation to affect [5] and exercise participation broadly [6], but it remains unclear the extent to which motivations may be associated with maladaptive exercise. The current study examined the relationships between exercise frequency and exercise motivations as they related to healthy exercise (i.e., physical activity performed for enjoyment and/or health maintenance) [7] and maladaptive exercise (i.e., high exercise frequency combined with the inability to reduce or stop exercising) [4] among university students, an at-risk population for the development of disordered eating and maladaptive exercise [8].

Conceptualizations of maladaptive and healthy exercise

There are multiple definitions of maladaptive exercise, including compulsive exercise [4], excessive exercise [9], exercise addiction [10], exercise dependence [11], obligatory exercise [12], and compensatory exercise [13]. Each of these conceptualizations contains an “excessive” component, suggesting that increased exercise frequency is a key aspect of maladaptive exercise behavior. Exercise frequency is often included in measures of maladaptive exercise [9], however, frequency alone is insufficient to capture and define maladaptive exercise [14]. Importantly, exercise frequency combined with other cognitive factors (e.g., preoccupation with weight and shape, exercise preoccupation) may result in maladaptive exercise behavior [15]. The current study used compulsive exercise (i.e., ritualistic, persistent exercise aimed at avoiding/reducing negative affect and/or body image concerns) [4] to represent the broader construct of “maladaptive” exercise.

Healthy exercise is less frequently studied than “maladaptive” exercise, and definitions of healthy exercise vary by population (i.e., what is healthy for one group may not be healthy for others). While the behavioral patterns that define healthy exercise may differ between individuals, cognitive patterns, such as enjoyment when exercising and wanting to exercise to maintain health, are typically used to characterize exercise as healthy [7]. Additionally, healthy exercise is often considered the opposite of “maladaptive” exercise and is not compulsive in nature.

Self-determination theory

Self-determination theory [16] posits that individuals who are internally motivated to perform a given behavior (i.e., intrinsic motivation) are more likely to perform that behavior, while individuals who are motivated by external factors (i.e., extrinsic motivation) may be less likely to engage in a particular behavior. Additionally, self-determination theory posits that intrinsic motivation is achieved when an individual satisfies their needs for autonomy, competence, and relatedness to other individuals [16]. Individuals’ need for self-growth is believed to contribute to and drive behavior, and internal sources of motivation are crucial with regard to behavioral outcomes. In general, individuals who are intrinsically motivated engage in activities because they provide satisfaction and pleasure, while individuals who are extrinsically motivated perform activities to obtain a reward unrelated to the activity itself (e.g., tangible rewards, punishment avoidance, recognition, approval) [17]. With regard to exercise and physical activity, intrinsic motivation is associated with exercise performed for enjoyment, while extrinsic motivation for exercise includes exercise performed for external rewards, such as achieving a thinner body or being recognized as an exerciser by one’s peers [17]. Intrinsic motivation is crucial for continued adherence to health-focused exercise programs [18] and is often a focus of health promotion programs [19]. While health-related motivations may be categorized as both intrinsic and extrinsic motivators [6], there is evidence that health maintenance motivation is more strongly associated with intrinsic than extrinsic motivation [20]. Importantly, intrinsic and extrinsic motivations are not mutually exclusive; individuals can report exercising for both intrinsic and extrinsic reasons.

Consistent with self-determination theory [16], intrinsic motivations for exercise are related to positive exercise outcomes (e.g., greater physical self-worth, decreased social physique anxiety) [21, 22], and extrinsic motivations for exercise are associated with greater general negative affect in non-clinical samples [5]. Findings from clinical samples are similar; feelings of autonomy, competence, and relatedness to others can buffer against sociocultural pressures that perpetuate a thin body ideal [23], and extrinsic motivation for exercise contributes to the development of compulsive exercise [24]. Conclusions about how exercise motivations predict exercise behaviors, however, are somewhat mixed. Some studies demonstrate that intrinsic, but not extrinsic, exercise motivations are associated with greater exercise frequency [6], while others report that both types of motivation are associated with greater exercise frequency [5]. A possible explanation for these equivocal findings is that exercise motivations research commonly focuses on either intrinsic or extrinsic exercise motivations. Importantly, because of their differing relationships with positive and negative cognitive outcomes (i.e., intrinsic motivation predicting greater self-worth, extrinsic motivation predicting negative affect), it could be that intrinsic and extrinsic exercise motivation independently predict whether exercise is healthy or compulsive. Because intrinsic and extrinsic motivations are independent, yet positively related constructs, whichever motivation is stronger could determine whether exercise is healthy or compulsive. Alternatively, it is possible that it is the combination of exercise frequency and either intrinsic or extrinsic motivation determines whether exercise behavior is healthy or compulsive. The present study extended the exercise motivations literature by simultaneously investigating intrinsic and extrinsic exercise motivations as predictors of exercise frequency among college students.

A second limitation of the exercise motivations literature is its primary outcomes of interest. First, the exercise motivations research primarily focuses on affective [5] and behavioral intention outcomes [6]. While exercise intentions are modestly correlated with actual behavior [25], it remains unclear how intrinsic and extrinsic motivations are associated with actual exercise behavior. Second, when exercise behavior is specified as an outcome, the extent to which the behavior may be healthy or unhealthy is not measured [e.g., 6]. The present study addresses these limitations by including both healthy and compulsive exercise as outcomes.

Current study

Established outcomes of exercise motivations include affect [5] and exercise intentions [6]. Yet, it is unknown how these motivations are associated with healthy and/or compulsive exercise specifically. As such, the current study investigated extrinsic and intrinsic exercise motivations as independent moderators of the associations between exercise frequency and 1) compulsive exercise and 2) healthy exercise. In line with self-determination theory [16], it was hypothesized that among individuals reporting greater intrinsic motivation, greater exercise frequency would be related to more healthy and less compulsive exercise. Conversely, we hypothesized that among individuals reporting higher extrinsic motivation, greater frequency would be related to more problematic and less healthy exercise.

Method

Participants and procedure

The current study was a secondary analysis of cross-sectional data collected from 699 participants from a large Southeastern university via an online Qualtrics survey. Participants were recruited using convenience sampling from the university’s research participant pool from the summer of 2020 to the spring of 2021. Inclusion criteria were that the individual was 1) at least 18 years old, and 2) currently enrolled in the university’s research participant pool. After providing informed consent, participants anonymously completed measures and received partial course credit upon completion. The study was deemed exempt by the university’s Institutional Review Board.

During data cleaning, 163 participants were removed for providing no data after giving informed consent and 90 participants were removed for missing at least two out of three attention checks [26]. After data cleaning, 446 individuals (50.9% female) were available for data analysis. The sample was primarily non-Hispanic (78.9%; White: 64.5%; Asian: 14.8%; Black or African American: 10.2%; Multiracial: 8.0%; Other: 2.3%; American Indian or Alaska Native: 0.3%). Hispanic individuals primarily identified as White (76.6%; Other: 11.7%; Black or African American: 4.3%; Multiracial: 3.2%; American Indian or Alaska Native: 3.2%; Asian: 1.1%). The average age of the sample was 20.10 (SD = 3.98).

Measures

Data were collected as part of a larger investigation of exercise behaviors, exercise identity, and disordered eating. The following measures are relevant to the current study.

Demographic information

Participants self-reported, age, sex, race, and ethnicity.

Exercise frequency

A multi-part item measuring exercise frequency was developed for the present study. Participants were asked how often they engaged in each of a number of types of exercise, on average, over the past four weeks. Multiple examples of each type of exercise were provided. Participants reported on the following exercise domains: cardiovascular (e.g., running), mixed modal training (e.g., Crossfit), weight or strength training (e.g., weightlifting), balance and stability training (e.g., yoga), and other (i.e., anything else done for physical activity). Response options were: 0 times, 1–2 times, 3–4 times, 5–6 times, and 7 + times. A sum score of exercise across the five domains was calculated as a measure of exercise frequency.

Exercise motivation

The 20-item goal content for exercise questionnaire measures the reasons for which individuals exercise [20]. This measure was developed within a self-determination theory framework and includes validated subscales that can be combined to measure intrinsic and extrinsic motivation. Items are rated on a scale from 1 (not at all important) to 7 (extremely important). The Social Affiliation (α = 0.87), Skill Development (α = 0.92), and Health Management subscales (α = 0.86) can be combined as a measure of Intrinsic Motivation (α = 0.91), while the Image (α = 0.91) and Social Recognition (α = 0.89) subscales, can be combined into a measure of Extrinsic Motivation (α = 0.91). The Intrinsic and Extrinsic subscales were used in the current study. Higher scores indicate a greater endorsement of the exercise motivation.

Exercise behaviors

The 18-item exercise and eating disorder questionnaire measures aspects of both compulsive and healthy exercise, particularly as they relate to eating behavior, body image, and broader health [7]. Items are rated on a scale from 0 (never) to 5 (always). Importantly, this measure has been validated in non-clinical male and female populations [7, 27]. The current study used the compulsive (α = 0.84) and positive and healthy exercise subscales (α = 0.78). Positive statements are reverse scored on this measure; however, for ease of interpretation, items were left untransformed when computing the positive and healthy exercise subscale so that higher scores would indicate more exercise, regardless of type of exercise.

Data analysis

Missing data were minimal (n = 5, 1.1% of all data) and listwise deletion was used. Assumptions of normality were upheld, as the skewness and kurtosis values of the dependent variables fell between − 1 and 1. All analyses were conducted in SPSS v26. Intrinsic and extrinsic motivations for exercise were entered as simultaneous independent moderators of the associations between exercise frequency and (1) compulsive exercise and (2) healthy exercise in multiple linear regressions using model 2 in Process v3.4 [28]. In cases where only one interaction term was statistically significant, a reduced model was retested using model 1 in Process v3.4 [28] with a single interaction term (motivation*exercise frequency) and a main effect of the other motivation. Interactions significant at the p < 0.05 level were probed at ± 1 standard deviation from the mean. When neither interaction term was statistically significant, reduced main effects models were tested using multiple linear regression. Unadjusted models (i.e., models including no covariates) were used to tested hypotheses.

Results

Correlations between variables of interest can be found in Table 1. As expected, all study variables were significantly and positively correlated. Based on the proposed severity thresholds [7], 22.3% of participants reported no symptoms of compulsive exercise, 22.8% fell within the low severity range, 38.9% fell within the moderate severity range, and 16% fell within the high severity range.

Table 1 Descriptive statistics and correlations of study variables (N = 441)
Full size table

Compulsive exercise

No moderation effects were observed for compulsive exercise, therefore, a main effects model was tested. Greater exercise frequency (b = 0.04, p < 0.01, f2 = 0.03) and stronger extrinsic exercise motivation (b = 0.27, p < 0.01, f2 = 0.14) were associated with more compulsive exercise, and these were small to medium effects. Intrinsic exercise motivation was not significantly associated with compulsive exercise (p = 0.93; see Table 2).

Table 2 Final regression models for each dependent variable (N = 441)
Full size table

Healthy exercise

Intrinsic exercise motivation, but not extrinsic motivation, moderated the association between exercise frequency and healthy exercise. Intrinsic exercise motivation remained a significant moderator of small effect in a reduced model excluding the extrinsic motivation interaction (b = − 0.02, p = 0.03, f2 = 0.01; see Table 2). Among individuals with lower intrinsic exercise motivation, greater exercise frequency was associated with more healthy exercise (b = 0.15, p =  < 0.01; see Fig. 1). Greater exercise frequency also was associated with more healthy exercise among individuals with high intrinsic motivation, but the association was weaker (b = 0.09, p =  < 0.01). Extrinsic exercise motivation was not related to healthy exercise (p = 0.06).

Fig. 1
figure 1

Intrinsic exercise motivation moderating the association between exercise frequency and positive and healthy exercise. Note. Both simple slopes are significant at the p < 0.01 level. Ex Exercise, Freq Frequency, Motiv Motivation, Low and High refer to ± 1 SD from the mean for the respective variable

Full size image

Discussion

Partially supporting self-determination theory [16] and hypotheses, extrinsic exercise motivation was only associated with compulsive exercise, and intrinsic exercise motivation moderated the relationship between exercise frequency and healthy exercise. Importantly, extrinsic motivation was associated with compulsive exercise independent of the frequency of exercise behaviors. This suggests that measuring multiple exercise motivations may be more important than capturing frequency when investigating compulsive exercise. Unexpectedly, while individuals with higher intrinsic motivation reported more healthy exercise compared to those with lower intrinsic motivation, the association between exercise frequency and healthy exercise was stronger for individuals with lower intrinsic motivation. Nonetheless, findings are consistent with the tenets of self-determination theory in that increasing individuals’ intrinsic motivation for exercise may lead to engagement in healthy, but not compulsive, exercise.

In line with previous research examining exercise motivation and affect [5], the present study found that extrinsic exercise motivation was associated with compulsive exercise, independent of exercise frequency. This extends the existing literature that primarily examines affective outcomes (e.g., negative affect) [5] by demonstrating that extrinsic exercise motivations may influence negative behavioral outcomes as well. The notion that greater extrinsic exercise motivation is associated with greater compulsive exercise also fits well within self-determination theory [16]. Compulsive exercise is strongly linked with external rewards, such as being viewed as thin or lean by one’s peers [29]. Thus, extrinsic exercise motivation may be a useful point of intervention among individuals engaging in compulsive exercise.

Contrary to the previous literature [e.g., 30], results suggest that the impact of intrinsic motivation on exercise behavior seems to depend on exercise frequency. This is in line with self-determination theory [16], as individuals who are exercising for intrinsic reasons are likely engaging in these behaviors in a more consistent way compared to individuals who are exercising with lower intrinsic motivation (i.e., the relationship is not as influenced by frequency). While lower intrinsic motivation may be associated with lower engagement in healthy exercise among individuals who exercise infrequently, intrinsic motivation was not associated with compulsive exercise. This suggests that interventions aimed at increasing intrinsic motivation among individuals who exercise infrequently may result in more healthy exercise within this population, without increasing the risk of engagement in compulsive exercise. To better understand the role of intrinsic exercise motivation on exercise behavior, additional factors, such as exercise identity [31] or peer norms [32, 33] should be investigated in attempts to better understand the purpose that exercise serves.

A positive association between intrinsic and extrinsic exercise motivations emerged, suggesting that these constructs are not orthogonal. Instead, these motivations may be best operationalized as two distinct, yet positively related, constructs. Similar to how the presence of positive affect does not necessarily indicate the absence of negative affect [34], the presence of intrinsic motivation does not indicate the absence of extrinsic motivation, and vice versa. It is likely that intrinsic and extrinsic exercise motivations exist on separate unipolar dimensions, as opposed to a single bipolar dimension. By conceptualizing and measuring these exercise motivations as distinct, yet related constructs, researchers will be better positioned to understand their unique, and potentially complex (e.g., high intrinsic and extrinsic motivation), influences on exercise behavior.

Implications

Compulsive exercise is associated with a variety of negative psychological and physical outcomes [4], and exercise motivation may be an important target for preventative interventions focused on compulsive exercise behaviors. Increasing intrinsic motivation is associated with participation in health and wellness promotion programs [19], and increasing intrinsic motivation may be particularly important for infrequent exercisers who report lower levels of healthy exercise. Additionally, evidence from broader lifestyle interventions targeting exercise motivations suggests that shifting individuals’ motivations can lead to beneficial health outcomes (e.g., successful weight management) [35]. It is possible that changing individuals’ reasons for exercise could prevent engagement in compulsive exercise. Research should continue to identify which specific types of exercise motivation are most related to (1) healthy physical and positive mental health outcomes and (2) compulsive exercise and other negative mental health outcomes to inform health promotion and health risk prevention programs.

Study strengths

The present study had a number of strengths. First, both healthy and maladaptive (i.e., compulsive) exercise were investigated. Research on exercise rarely specifies whether the exercise behavior is healthy or compulsive, contributing to an assumption that all exercise is beneficial to health and wellbeing. Second, both intrinsic and extrinsic motivation were measured, and their simultaneous effects on both healthy and compulsive exercise were investigated. Exercise motivation research often tests the effects of only one type of motivation [e.g., 36], which may contribute to the belief that intrinsic and extrinsic exercise are orthogonal. Third, the sample was sex balanced. While there is limited evidence of sex differences in rates of compulsive exercise [37], many studies examining the influences of exercise motivations include all-female samples, limiting generalizations to the broader population of exercisers. Fourth, the present study used a non-clinical, young adult sample. To date, compulsive exercise primarily has been investigated in samples of individuals with eating disorders [e.g., 38]. The present findings provide important information about how exercise motivations are associated with healthy and compulsive exercise in non-clinical samples.

Limitations and future directions

Despite study strengths, a number of limitations merit consideration. First, while self-determination theory proposes a wide range and presentation of motivations [16], the current study simplified these motivations into broader categories of “intrinsic” and “extrinsic” motivation to reflect common practices of the existing literature [17, 30]. As this study was the first to test the associations between exercise motivations and negative behavioral outcomes, it was useful to conceptualize the motivations in a manner consistent with previous research. Future studies may benefit from investigating the specific subtypes of exercise motivations (e.g., exercising for body image reasons; exercising as a social experience), as different specific motivations may be differentially associated with healthy and compulsive exercise behavior. A second limitation is the self-report nature of exercise frequency. Retrospective reports of exercise behaviors may be biased such that exercise behaviors are overreported [39], especially when participants also are asked to report on their current exercise behavior. Participants may have overestimated their frequency of exercise behaviors, which could have potentially inflated the magnitudes of the observed relationships. Future research should utilize an objective measure of exercise frequency (e.g., Fitbit, Apple Watch, etc.) to investigate if exercise motivations are associated with objective reports of exercise frequency in addition to whether exercise behaviors are self-reported as healthy or compulsive. Related, the current study operationalized exercise frequency as the total self-reported exercise frequency across multiple domains. Future work may benefit from including measures of exercise intensity and duration, as they may influence the relationships between exercise motivation and healthy and compulsive exercise [40]. Additional factors that merit consideration include athlete status and competition level [41], exercise identity [31], current and past disordered eating behavior [36], and body mass index [42], as individuals who differ on these variables may report unique patterns of exercise motivation and compulsive exercise. Last, because the current data were cross-sectional, causal conclusions cannot be made. Future studies must assess the temporal sequencing of various exercise motivations, exercise frequency, and different types of exercise behavior (e.g., healthy versus compulsive) using longitudinal methodology or ecological momentary assessment. Indeed, individuals can engage in healthy and compulsive exercise during different exercise episodes [43] and exercise motivations are often treated as trait-level characteristics, but it is possible that the specific motivations for exercise fluctuate over time [30]. Fluctuations in types of exercise motivations may be important to whether exercise is healthy or compulsive during a given exercise episode.

Conclusion

Supporting conceptualizations of maladaptive exercise that consider both exercise frequency and cognitive factors (e.g., weight concerns), the current study demonstrated that exercise motivations account for significant variance in compulsive and healthy exercise in young adult college students. Consistent with extant work, extrinsic exercise motivations were associated with compulsive exercise, regardless of exercise frequency, while the association between intrinsic motivation and healthy exercise was dependent on exercise frequency. Decreasing extrinsic motivation and increasing intrinsic motivation may be important aims for prevention of compulsive exercise behavior. Findings underscore the relevance of exercise motivations to exercise behavior, as well as highlight the importance of continued investigations into how these motivations influence whether a person will exercise in a healthy or compulsive manner.

Strengths and limits

The primary strengths of the current study include the measurement of both adaptive and maladaptive exercise behaviors, the simultaneous inclusion of both intrinsic and extrinsic exercise motivations, and the use of a relatively sex-balanced non-clinical sample, which contributes to the generalization of findings to the broader population. The key limitations of the current study are its cross-sectional nature and simplified operationalization of exercise motivations as either intrinsic or extrinsic.

What is already known on this subject?

In general, intrinsic motivation for exercise is associated with positive outcomes (e.g., greater psychological wellbeing), while extrinsic motivation is related to negative outcomes (e.g., greater negative affect). Some studies have found that only intrinsic exercise motivation is associated with greater exercise frequency, while others have found that both types of motivation were associated with greater exercise frequency. It remains unknown whether intrinsic and extrinsic exercise motivations have differing relationships with healthy and compulsive exercise behaviors and whether motivation may change the association between exercise frequency and exercise behavior.

What this study adds?

Previous work primarily has examined affective outcomes of exercise motivations. The current study extended this work to both healthy and maladaptive behavioral outcomes and demonstrated that extrinsic motivation was uniquely associated with compulsive exercise, while level of intrinsic motivation changed the strength of the relationship between exercise frequency and healthy exercise. This study is also novel in that it assessed both intrinsic and extrinsic motivations for exercise simultaneously. Irrespective of exercise frequency, extrinsic exercise motivation may be a useful target for preventing compulsive exercise, while enhancing intrinsic exercise motivation may contribute to healthy exercise behaviors.