Replacing sedentary behavior with physical activity reduces symptoms of anxiety and depression: a study with young adults

Abstract

Background

Physical activity (PA) and sedentary behavior (SB) are movement behaviors that have been associated with mental health. Evidence suggests that replacing SB with PA may influence mental disorders. This study aimed to evaluate the effects of replacing time spent in SB with moderate-intensity PA (MPA) and vigorous-intensity (VPA) on the symptoms of anxiety and depression among college students.

Methods

This cross-sectional study used data from a multicenter survey conducted among undergraduate students from eight public universities in Brazil. Data were collected between October 2021 and February 2022 using an online questionnaire sent via email. The outcomes were anxiety and depression symptoms, assessed using the Depression, Anxiety, and Stress Scale-21. Exposure to SB was evaluated by the total sitting time and PA level, considering the type of exercise, weekly frequency, and duration in minutes. Subsequently, PA was classified according to its intensity. The explanatory variable was adherence to the 24-hour movement guidelines, assessed by combining the time spent in SB and PA. Based on this, three groups were formed: (1) complete adherence, (2) partial adherence, and (3) non-adherence. An isotemporal substitution logistic model was used to assess the effects of the different SB, MPA, and VPA periods on mental health.

Results

A total of 8,059 young adults participated in the study, with a mean age of 23.9 years (SD: ± 6.28). Most students reported spending more than 9 h/day in SB (55.09%), and 48.28% were physically inactive. The multivariate analysis revealed an association between non-adherence to movement guidelines and mental health (p < 0.001), showing a dose-response gradient. Students who did not adhere to the guidelines were more likely to have symptoms of anxiety [OR: 1.89 (95% CI: 1.67–2.14)] and depression [OR: 2.49 (95% CI: 2.19–2.82)]. Furthermore, in the isotemporal analysis, replacing SB with equivalent amounts of time in MPA and VPA reduced the odds of mental disorder symptoms in all models evaluated.

Conclusion

Replacing SB with equivalent quantities of MPA or VPA reduces the odds of symptoms of anxiety and depression. Therefore, it is essential to develop public policies that encourage increased PA levels and reduced SB in university environments to promote mental health and improve the physical health of students with mental disorder symptoms.

Introduction

Mental disorders, such as anxiety and depression, are considered important public health problems that pose considerable challenges to population health, especially among young adults aged 18–25 years [1, 2]. This period, marked by the transition from adolescence to adulthood, often coincides with the entry into university life. During this time, students are frequently exposed to stressors, such as academic pressures, new relationships, and difficulties managing time and finances. These factors may increase the risk of developing mental disorders [1, 3, 4].

Epidemiological studies have shown that the prevalence of mental disorders among students has increased in recent years [5–7], with a significant increase observed during the COVID-19 pandemic [8, 9]. Sudden changes in daily routines, combined with the demands of a new teaching and learning model — resulting from university closures and the transition from in-person to virtual instruction — have generated fear and uncertainty, significantly contributing to the worsening of mental health disorders, increased psychological distress, and intensified negative emotions among university students [10, 11].

Owing to the high prevalence and consequences of mental disorders in young adults, researchers have focused their efforts on developing actions and strategies to mitigate their negative impact on health [1, 2]. This has also driven interest in investigating the relationship between compliance with behavioral guidelines and mental disorders [12].

From a movement perspective, the 24 h of the day are divided into three types of behavior: sleep, sedentary behavior (SB)—defined as any activity with energy expenditure ≤ 1.5 metabolic equivalents, performed in a sitting, reclining, or lying posture [13]—, and physical activity (PA), which can range from light to moderate or vigorous intensities [14, 15]. Altogether, these behaviors, called to as “24-hour movement behaviors,” form a continuum that spans from no movement to intense movement [14, 16] and can have important implications for individuals’ physical and mental health and well-being [15, 16].

Sleep and PA are known to play protective roles in the prevention of mental disorders, contributing to the regulation of the hypothalamic-pituitary-adrenal (HPA) axis, the neuroendocrine system responsible for regulating stress in the body. In contrast, SB has been associated with an increased risk of anxiety and depression through several mechanisms, such as disturbances in biological functions, including central nervous system (CNS) excitation, direct cognitive effects, and reduced social interaction [8, 17, 18]. Additionally, prolonged SB may affect PA engagement, thereby limiting the well-established benefits of PA on physical and mental health [19].

In an academic context, these movement behaviors, especially PA and SB, are particularly affected because university routines often involve sedentary activities, such as sitting for long periods in classrooms or studying. Furthermore, factors such as lack of time, motivation, and psychological issues worsen this scenario among students, especially concerning the practice of PA [8, 20–22].

Until the middle of the second decade of the 21st century, around 2015, the implications of the time spent on each movement behavior for physical and mental health were assessed in isolation or with partial adjustments [14]. In this context, a new approach to public health promotion, which integrates all movement behaviors throughout the day, has be en increasingly explored in epidemiological studies [12, 14, 23], considering that the time spent on movement behaviors during the 24 h is intrinsically collinear and codependent [14]. Specifically, any increase in the time dedicated to a behavior throughout the day (e.g., SB) necessarily reduces the time available for other movement behaviors (e.g., sleep or PA) [17, 24]. Thus, from the perspective of movement, it is more interesting to consider the overall composition of time use in 24 h rather than analyzing the behaviors in isolation [24].

Research has investigated the association between the hypothetical effects of reallocating a given amount of time spent on one movement behavior to another and its impact on health outcomes using isotemporal substitution statistical modeling [14, 15, 24, 25]. Despite the growing interest in this topic, existing evidence has focused mainly on outcomes related to physical health, such as obesity and metabolic biomarkers [14, 15, 26]. Moreover, research examining the reallocation of time from SB to PA concerning health has been conducted in specific populations [15], such as children [26, 27], adolescents [27], adults, and older adults [25, 28]. However, there is a knowledge gap regarding studies conducted with young adults, the target audience of our study. As identified in the literature, the effect of isotemporal substitution of movement behaviors on the mental health of university students remains poorly explored [3, 29]. It is essential to evaluate the effect of time spent on movement behaviors to formulate public health recommendations considering the entire energy expenditure spectrum. Considering the aspects mentioned above, this study aimed to evaluate the effects of replacing time spent in SB with moderate-intensity PA (MPA) and vigorous-intensity (VPA) on symptoms of anxiety and depression among college students. We hypothesized that reallocating time from SB to an equivalent duration of PA would reduce the likelihood of young adults presenting symptoms of these mental disorders, whereas the inverse substitution (i.e., replacing MPA or VPA with SB) would increase this probability.

Methods

A cross-sectional study was conducted with data from the survey “Symptoms of anxiety disorder and depression among university students in Minas Gerais: a multicenter study (Project on Anxiety and Depression in University Students, PADu-multicenter)” between October 2021 and February 2022. This was conducted with a sample of students enrolled in undergraduate courses at eight public universities in Brazil, regardless of the academic period.

The participating universities were: Federal University of Ouro Preto (UFOP), in partnership with the Federal University of Minas Gerais (UFMG), Federal University of São João del-Rei (UFSJ), Federal University of the Jequitinhonha and Mucuri Valleys (UFVJM), Federal University of Juiz de Fora (UFJF), Federal University of Uberlândia (UFU), Federal University of Lavras (UFLA) and Federal University of Alfenas (UNIFAL-MG). The study was approved by the Research Ethics Committee of the coordinating center (protocol number 43027421.3.1001.5150) and by the Research Ethics Committees of all universities involved.

For the PADu-multicenter sample, 118,828 undergraduate students with active registration at universities were considered eligible. Of these, 8,650 completed the full questionnaire, resulting in a response rate of 7.3%, which is consistent with those reported in similar studies in the literature. The post hoc power analysis demonstrated that the study had sufficient statistical power (greater than 90%) to detect meaningful differences and associations, thereby supporting the robustness and validity of the findings.

The PADu-multicenter study included students of both sexes aged 18 years or older. Those who did not complete the questionnaire, postgraduate students, residents, and students who were away from academic activities or were on exchange during data collection were excluded from the study.

Students who did not answer all questions on the explanatory variable were excluded, resulting in a final sample of 8,059 participants. We did not perform any a priori sample calculations. Instead, we used a sample of students who had participated in the survey. A posteriori sample power calculation indicated that the study had sufficient power (> 90%) to verify significant differences and associations, thus ensuring the robustness and reliability of the results.

Students were recruited by emailing an informative invitation regarding the study and a link to the questionnaire. Additionally, as part of the communication, awareness-raising, and recruitment strategy, the PADu-multicenter study was widely publicized on social media and universities’ official websites, in addition to being promoted in laboratories, academic centers and directories, tutoring programs, and study and research groups.

A self-administered questionnaire structured into thematic blocks on socio-demographic and academic characteristics, lifestyle habits, and health conditions was used to collect the data available on Google Forms. Students were invited to participate in the study voluntarily, and participation began with access to the questionnaire, which was preceded by agreement with the Free and Informed Consent Form, presented electronically and available for download.

For more information on the methodology of the multicenter PADu study, see the previous publication by Barbosa et al. [30]

Study variables

Outcome variables: symptoms of anxiety and depression

The outcomes were symptoms of anxiety and depression assessed using the Depression Anxiety Stress Scale-21 (DASS-21), an instrument translated and validated in Portuguese by Vignola and Tucci to ensure its relevance and effectiveness in the Brazilian cultural context. This scale comprises three subscales, each with seven items that self-report the individual’s symptoms in the week before completing the questionnaire [31].

In the psychometric validation of the Brazilian version of the DASS-21, strong correlations were observed among the items of each subscale, with Cronbach’s alpha coefficients of 0.92 for depression, 0.90 for stress, and 0.86 for anxiety, indicating good internal consistency for each subscale [31]. The findings by Vignola and Tucci (2014) confirm the reliability and validity of the Brazilian Portuguese version of the DASS-21, demonstrating that each subscale adequately assesses its respective construct. This supports the instrument’s ability to measure emotional states separately, legitimizing the independent use of each subscale.

Responses to the DASS-21 items were scored on a four-point Likert-type scale ranging from 0 (not applicable at all) to 3 (applicable very much or most of the time). The final result is calculated by adding the scores of the items of each subscale and multiplying by two to correspond to the score of the original scale (DASS-42) [31, 32]. From this final score, cut-off points are established that allow the classification of symptoms of anxiety, depression, and stress into levels such as “normal,” “mild,” “moderate,” “severe,” and “extremely severe.” [31].

In this study, only the symptoms of anxiety and depression were analyzed, which were subsequently categorized as the absence of symptoms “no” (normal and mild) and the presence “yes” (moderate to extremely severe).

Explanatory variable: movement behaviors

Leisure-time physical activity

Leisure-time PA was assessed based on three questions adapted from the Surveillance System for Risk and Protection Factors for Chronic Diseases by Telephone Survey (Vigitel), a population-based survey conducted by the Brazilian Ministry of Health [33, 34]. The Vigitel questionnaire was validated and proven to be reliable and comparable to the Global Physical Activity Questionnaire (GPAQ), the reference method recommended by the World Health Organization (WHO) for measuring PA. The questionnaire used by Vigitel appears to be reliable for monitoring trends in PA indicators during leisure time, the domain assessed in the present study, and is comparable to the GPAQ in most aspects of PA assessment [35].

First, the participants were asked whether they had practiced any PA. Students who responded positively were invited to detail the type of PA they practiced and could choose more than one modality. For each activity, participants reported their weekly frequency and duration in minutes. The practice of PA was assessed based on a list of 24 types of physical exercise or sports categorized according to intensity, which was determined using the Compendium of PA Codes and MET Intensities, classifying the activities as moderate or vigorous PA.

Treadmill walking, hiking, weight training, cycling, stretching, Pilates, yoga, swimming, water aerobics, fighting and martial arts (jiu-jitsu, karate, judo, boxing, Muay Thai, capoeira), cycling (including stationary cycling), volleyball or footvolley, and dancing (ballet, ballroom dancing, belly dancing, forró, and axé) were activities classified as moderate-intensity practices. In contrast, vigorous-intensity activities include running, treadmill running, functional training, CrossFit, aerobic gymnastics (spinning, step, and jump), tennis, basketball, soccer or futsal, high-intensity interval training, and team sports [33, 36].

To determine the variable “PA during leisure time,” the frequency of each activity evaluated was multiplied by the time in minutes, resulting in the total time spent on physical activities during leisure time. Students who practiced moderate-intensity PA for at least 150 min per week or 75 min of vigorous-intensity PA were classified as “physically active during leisure time.” Conversely, those who did not practice PA or did not meet these recommendations were classified as “physically inactive during their leisure time.” [37].

Sedentary behavior

SB was measured by total time spent sitting during the week and on weekends, based on a question adapted from a similar question in the long version of the International Physical Activity Questionnaire (IPAQ), which is validated to assess the level of PA and SB: [38] “Currently, how much time (in hours) on average do you spend sitting (include time spent on cell phone, TV, computer, tablet, books, car and bus) per day?”.

For analysis purposes, the weighted average of SB was calculated by adding the time spent in SB during the week (multiplied by 5) and on weekends (multiplied by 2), then dividing the result by 7 to represent the daily average [39].

Subsequently, SB was classified based on a cut-off point of ≥ 9 h/day (< 9 h and ≥ 9 h). This classification is based on a meta-analysis that covered more than 1 million participants from 19 different studies, indicating that individuals with high levels of SB have a higher risk of all-cause mortality. [40] It is noteworthy that there is no consensus in the literature on a cut-off point for SB in adults, especially concerning mental health outcomes such as anxiety and depression. In this study, as in that of Barbosa et al. [39] we chose to adopt the cut-off point established by Ku et al. [40] given its wide acceptance in the existing literature [39, 41] and the consistency it has demonstrated concerning other physical health outcomes, including all-cause mortality and cardiovascular diseases.

Combinations of movement behaviors

Adherence to the 24-hour movement guidelines, defined as an explanatory variable, was assessed based on the combination of time spent on PA and SB, resulting in three groups: (1) full adherence: students with less than 9 h/day of SB and physically active; (2) partial adherence: students with high SB (≥ 9 h/day) or inactive; (3) non-adherence: students with more than 9 h/day of SB and inactive. It is essential to highlight that the present study did not consider all movement-related behaviors because sleep was not assessed owing to the absence of a specific question in the questionnaire.

Covariates

The covariates used to describe the sample and assess the relationship between symptoms of anxiety and depressive were grouped based on socio-demographic characteristics, academic aspects, and health conditions. The socio-demographic covariates were age (18–20 years, 21–22 years, 23–25 years and ≥ 26 years), biological sex (men and women), race or skin color (white, brown, black and yellow or Indigenous or other), sexual orientation (heterosexual, homosexual, bisexual and asexual or other), marital status (single, married or stable union and widowed or divorced), housing (with family members and without family members), total family income (≤ 1–2 minimum wages, 3–5 minimum wages 6–10 minimum wages and > 10 minimum wages) and education level of the head of the family (no education or incomplete elementary education, complete elementary education or incomplete high school, complete high school or incomplete higher education, complete higher education). In the academic domain, the variable assessed was the area of ​​knowledge of the course (life sciences, exact sciences, humanities, social sciences, and applied sciences) in which students were enrolled at the time of the survey.

As a health condition, body mass index (BMI) was assessed and calculated based on weight (kg) and height (m²) self-reported by the participants. The BMI was classified according to the values ​​the WHO recommended for adolescents [42], adults [43], and older adults [44]. Individuals classified as underweight and eutrophic were grouped in the “not overweight” category, while those classified as overweight and obese in the “overweight” category [45]. In the block on health condition, the variables of medical diagnosis of anxiety and depression were also assessed, obtained through the question: “Has a doctor or other health professional ever told you that you have: anxiety or depression?” The answers were categorized as “no” or “yes.”

Statistical analysis

The analyses were conducted using the statistical software Stata version 13.0 (Stata Corporation, College Station, USA). To describe the sample, the variables were analyzed descriptively using relative frequencies and 95% confidence intervals (95% CI). Further, Pearson’s square test was used to assess the relationship between mental health and covariates.

Univariate and multivariate binary logistic regression analyses assessed the association between movement behaviors and mental health. A theoretical causality model based on a Directed Acyclic Graph (DAG) was developed to select appropriate adjustment variables using the online software Dagitty version 3.2. Outcome variables (symptoms of anxiety and depression), exposure (movement behaviors), and covariates were considered (Fig. 1). A backdoor criterion defines the minimum set of variables to be included in the analyses to avoid unnecessary adjustments, spurious associations, and estimation errors. The multivariate model was adjusted for biological sex, age, race or skin color, marital status, housing, education level of the head of the family, area of knowledge, and nutritional status. In the multivariate logistic models, we decided to include only the variable education of the household head without adjusting for family income. This choice was justified by the collinearity between the variables and considering that, in the academic context, parental education may have a more relevant influence on students’ permanence at university, reducing dropout rates.

Fig. 1.

DAG of the association between movement behaviors and mental health. Legend: DAG: Directed acyclic graph. The variable in green and with the “►” symbol inside the rectangle was the explanatory variable; in blue and with the letter “I” inside is the outcome variable, composed of symptoms of anxiety and depression. The figure shows only the variables that were selected for multivariate. The variables are: biological sex, age, race or skin color, marital status, housing, education level of the head of the family, area of knowledge and nutritional status. The arrows indicate the causal relationships between the variables

An isotemporal substitution approach was used to verify the hypothetical effects of the reallocation of time spent in SB, MPA, and VPA on mental health. Using a regression model, this statistical method mathematically estimates the associations of substituting one type of movement behavior with another of equal time, keeping the total time of the day constant [46].

For analytical purposes, isotemporal substitution models were used to evaluate the effect of reallocating 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, and 60 min/day spent in SB, MPA, and VPA on anxiety and depression symptoms. Initially, the total time variable was calculated by adding the minutes dedicated to each movement behavior (total time = SB + MPA + VPA) to provide an overview of the total time spent. Then, all-time variables were normalized by their respective fractions; the variables were divided by the time interval that would be reallocated. For example, when analyzing a 10 min replacement, each variable was divided by 10. Similarly, if the replacement time is 15 min, it is divided by 15. In the regression model, all movement behavior variables, except the one being replaced (SB, MPA, or VPA), were included along with the total time variable. Covariates of interest were also included in the model. Isotemporal replacement analyses were performed by estimating the odds ratios (OR) with a 95% CI.

Results

A total of 8,059 students were evaluated, most of whom were women, self-identified as white, heterosexual, single, and living with their families. The average age of the students was 23.9 ± 6.28 years. Regarding the socioeconomic status of the participants’ families, 39.5% reported that the head of the family had completed higher education, and 40.8% reported a family income between 3 and 5 minimum wages. In the academic domain, students reported being enrolled in the following areas of knowledge: 39.9% in exact sciences, 31.4% in life sciences, and 28.7% in human, social, and applied sciences. Regarding health status, 34.6% of students were classified as overweight, 56.4% reported a medical diagnosis of anxiety, and 30.7% reported depression (Table 1).

Table 1.

Sociodemographic characteristics, academic aspect, and health conditions according to anxious and depressive symptomatology in young adults, PADu-multicentric study (2021/2022), n = 8,059

Variables	Prevalence % (95% CI)	Symptoms of anxiety % (95% CI)		P value*	Symptoms of depression % (95% CI)		P value*
		Absence (40.6%)	Presence (59.4%)		Absence (37.3%)	Presence (62.7%)
Biological sex				< 0.001			< 0.001
Men	34.6 (33.6–35.7)	45.3 (43.6–47.0)	27.4 (26.1–28.6)		38.9 (37.2–40.7)	32.1 (30.8–33.4)
Women	65.4 (64.3–66.4)	54.7 (53.0-56.4)	72.6 (71.4–73.9)		61.1 (59.3–62.8)	67.9 (66.6–69.2)
Age				< 0.001
18–20 years	29.5 (28.6–30.6)	27.6 (26.1–29.2)	30.9 (29.6–32.2)		30.4 (28.8–32.0)	29.0 (27.8–30.3)
21–22 years	24.7 (23.8–25.6)	24.6 (23.2–26.1)	24.7 (23.5–26.0)		24.8 (23.2–26.3)	24.7 (23.5–25.9)
23–25 years	23.1 (22.2–24.0)	21.8 (20.4–23.2)	24.0 (22.8–25.2)		21.0 (19.6–22.5)	24.3 (23.2–25.5)
≥ 26 years	22.7 (21.8–23.6)	26.0 (24.5–27.5)	20.4 (19.3–21.6)		23.9 (22.4–25.5)	22.0 (20.9–23.1)
Race or skin color				0.012			< 0.001
White	55.4 (54.3–56.5)	57.1 (55.4–58.8)	54.2 (52.8–55.6)		58.5 (56.7–60.3)	53.5 (52.1–54.9)
Brown	31.0 (30.0–32.0)	30.4 (28.8–32.0)	31.4 (30.1–32.8)		30.4 (28.8–32.1)	31.3 (30.1–32.6)
Black	12.2 (11.5–12.9)	10.9 (9.9–12.1)	13.0 (12.1–14.0)		9.7 (8.7–10.8)	13.7 (12.7–14.7)
Yellow or Indigenous or other	1.5 (1.2–1.7)	1.6 (1.2–2.1)	1.4 (1.1–1.7)		1.4 (1.0-1.9)	1.5 (1.2–1.9)
Sexual orientation				< 0.001			< 0.001
Heterosexual	68.2 (67.1–69.2)	77.1 (75.6–78.5)	62.1 (60.7–63.5)		77.3 (75.8–78.8)	62.7 (61.3–64.0)
Homosexual	9.1 (8.5–9.8)	7.6 (6.8–8.6)	10.2 (9.3–11.1)		7.0 (6.2-8.0)	10.4 (9.6–11.3)
Bisexual	20.1 (19.2–21.0)	13.9 (12.7–15.1)	24.4 (23.2–25.7)		14.6 (13.3–15.9)	23.5 (22.3–24.7)
Asexual or other	2.6 (2.2–2.9)	1.4 (1.1–1.9)	3.3 (2.9–3.9)		1.1 (0.7–1.5)	3.5 (3.0–4.0)
Marital status				< 0.001			< 0.001
Single	90.7 (90.1–91.4)	88.6 (87.4–89.6)	92.2 (91.4–93.0)		88.5 (87.3–89.6)	92.1 (91.3–92.8)
Married or stable union	8.2(7.6–8.8)	10.2 (9.2–11.3)	6.8 (6.1–7.5)		10.0 (9.0-11.2)	7.0 (6.4–7.8)
Widowed or divorced	1.1 (0.9–1.4)	1.3 (0.9–1.7)	1.0 (0.7–1.3)		1.4 (1.1–1.9)	0.9 (0.07–1.2)
Housing				0.051			0.299
With family members	76.4 (75.5–77.3)	77.5 (76.0-78.9)	75.6 (74.4–76.8)		77.0 (75.5–78.5)	76.0 (74.8–77.2)
Without family members	23.6 (22.7–24.5)	22.5 (21.1–24.0)	24.4 (23.2–25.6)		23.0 (21.5–24.5)	24.0 (22.8–25.2)
Education level of the head of the family				< 0.001			< 0.001
No education or incomplete elementary education	15.2 (14.4–16.0)	12.3 (11.3–13.5)	17.1 (16.1–18.2)		12.8 (11.6–14.0)	16.6 (15.6–17.7)
Complete elementary education or incomplete high school	11.0 (10.3–11.7)	10.9 (9.9–12.0)	11.1 (10.2–12.0)		10.9 (9.8–12.1)	11.0 (10.2–11.9)
Complete high school or incomplete higher education	34.3 (33.3–35.4)	33.9 (32.3–35.5)	34.6 (33.3–36.0)		32.5 (30.8–34.2)	35.4 (34.1–36.7)
Complete higher education	39.5 (38.4–40.6)	42.9 (41.2–44.6)	37.2 (35.6–38.6)		43.8 (42.1–45.6)	36.9 (35.6–38.3)
Total family income a				< 0.001			< 0.001
≤ 1–2 minimum wages	32.1 (31.1–33.2)	25.1 (23.6–26.6)	37.0 (35.6–38.4)		25.7 (24.1–27.3)	36.0 (34.6–37.3)
3–5 minimum wages	40.8 (39.7–42.0)	42.4 (40.7–44.2)	39.8 (38.3–41.2)		40.8 (39.0-42.7)	40.9 (39.5–42.3)
6–10 minimum wages	17.2 (16.4–18.1)	19.6 (18.3–21.1)	15.6 (14.6–16.7)		19.9 (18.4–21.4)	15.7 (14.7–16.8)
> 10 minimum wages	9.8 (9.1–10.5)	12.9 (11.7–14.1)	7.7 (6.9–8.5)		13.6 (12.4–15.0)	7.5 (6.8–8.3)
Knowledge of the course				< 0.001			< 0.001
Exact sciences	39.9 (38.9–41.0)	42.9 (41.2–44.6)	37.9 (36.5–39.3)		41.2 (39.5–43.0)	39.1 (37.8–40.5)
Life sciences	31.4 (30.4–32.4)	33.0 (31.4–34.7)	30.3 (29.0-31.6)		33.9 (32.3–35.6)	29.9 (28.6–31.1)
Humanities, social sciences, and applied sciences	28.7 (27.7–29.7)	24.1 (22.7–25.6)	31.8 (30.5–33.2)		24.9 (23.3–26.4)	31.0 (29.7–32.3)
Nutritional status b				0.002			< 0.001
Not overweight	65.4 (64.3–66.4)	67.4 (65.8–69.0)	64.0 (62.6–65.3)		69.1 (67.4–70.7)	63.2 (61.8–64.5)
Overweight	34.6 (33.6–35.7)	32.6 (31.0-34.2)	36.0 (34.7–37.4)		30.9 (29.3–32.6)	36.8 (35.5–38.2)
Medical diagnosis of anxiety				< 0.001			< 0.001
No	43.6 (42.6–44.7)	62.5 (60.8–64.1)	30.8 (29.5–32.1)		57.2 (55.4–59.0)	35.6 (34.3–36.9)
Yes	56.4 (55.3–57.4)	37.5 (35.9–39.2)	69.2 (67.9–70.5)		42.8 (41.0-44.6)	64.4 (63.1–65.7)
Medical diagnosis of depression				< 0.001			< 0.001
No	69.3 (68.3–70.3)	82.5 (81.1–83.7)	60.3 (58.9–61.7)		86.1 (84.8–87.3)	59.3 (57.9–60.6)
Yes	30.7 (29.7–31.7)	17.5 (16.3–18.9)	39.7 (38.3–41.1)		13.9 (12.7–15.2)	40.7 (39.4–42.1)

Note:

*P value obtained using bivariate logistic regression; In bold: the statistically significant variables in the bivariate analysis

^a Minimum wage in force in Brazil in 2021 = R$1.100.00 (approximately $223.22);

^b Nutritional status: Individuals classified as underweight and eutrophic were grouped in the “not overweight” category, while those classified as overweight and obese in the “overweight” category

Through multivariable logistic regression analysis between movement behaviors and the presence of anxious and depressive symptoms, it was observed that students with SB of 9 h/day or more had greater chances of anxiety symptoms (OR: 1.37 ([95% CI: 1.24–1.50]) and depression symptoms (OR: 1.61 [95% CI: 1.47–1.77]) when compared to students with less than 9 h of SB. When evaluating the practice of PA, the association demonstrated that the chance of presenting symptoms of anxiety (OR: 1.52 (95% CI: 1.38–1.67]) and symptoms of depression (OR: 1.76 [95% CI: 1.60–1.93]) was greater in students who did not practice or did not meet the PA recommendations than in physically active students. Furthermore, when combined, a dose-response gradient was observed in the association between adherence to movement guidelines and mental health. Students who did not adhere to the guidelines were more likely to experience symptoms of anxiety (OR: 1.89 [95% CI: 1.67–2.14]) and symptoms of depression (OR: 2.49 [95% CI: 2.19–2.82]) compared to individuals who fully adhered to the guidelines (less than 9 h of SB per day and were physically active) (Table 2).

Table 2.

Association of movement behaviors with anxiety and depression symptoms, PADu-multicenter (2021–2022)

Movement behaviors	Prevalence % (95% CI)	Symptoms of anxiety			Symptoms of depression
		Univariate OR (95% CI)	Multivariate OR (95% CI)	P value*	Univariate OR (95% CI)	Multivariate OR (95% CI)	P value*
Sedentary behavior
< 9 h/day	44.90 (43.82–45.99)	1	1		1	1
≥ 9 h/day	55.09 (54.00−56.18)	1.35 (1.23–1.48)	1.37 (1.24–1.50)	< 0.001	1.60 (1.46–1.75)	1.61 (1.47–1.77)	< 0.001
Physical activity 1
Active	51.72 (50.63–52.81)	1	1		1	1
Inactive	48.28 (47.19–49.37)	1.57 (1.44–1.72)	1.52 (1.38–1.67)	< 0.001	1.79 (1.63–1.96)	1.76 (1.60–1.93)	< 0.001
Adherence to wake-based movement guidelines 2
Complete adherence	26.60 (25.65–27.58)	1	1		1	1
Partial adherence	43.42 (42.3–44.50)	1.45 (1.30–1.62)	1.44 (1.28–1.61)	< 0.001	1.65 (1.48–1.84)	1.65 (1.47–1.84)	< 0.001
Non-adherence	29.98 (29.0−30.99)	1.95 (1.73–2.19)	1.89 (1.67–2.14)	< 0.001	2.55 (2.25–2.88)	2.49 (2.19–2.82)	< 0.001

Note: CI: confidence intervals; OR: odds ratio; SB: Sedentary behavior; PI: physically inactive

¹ PI: < 150 min/week of moderate physical activity or < 75 min/week of vigorous activity

² Adherence to movement guidelines: Complete adherence (i.e. individuals with less than 9 h of SB per day and no PI), partial adherence (SB elevated or PI), and non-adherence (SB elevated and PI)

Bold values ​​indicate statistical significance (p-value < 0.05)

Multivariate logistic regression adjusted for sex, age, skin color, residence, education of the head of the family and area of ​​knowledge

Isotemporal substitution models indicated that increasing the time spent on PA and VPA while reducing the time spent on SB decreased the odds of students experiencing symptoms of anxiety and depression. There was a significant dose-response relationship in which greater time substitutions resulted in a more pronounced protective effect (p < 0.05). When comparing the practice of MPA with that of VPA, it was found that replacing moderate activity with vigorous activity reduced the chances of anxiety symptoms (p < 0.05). However, an inverse association was observed when VPA was replaced with MPA. Regarding depression symptoms, the replacement of PA, regardless of intensity, resulted in the same effect in all models (p > 0.05). Tables 3 and 4 presents the results of the isotemporal models.

Table 3.

Isotemporal substitution models of movement behaviors in anxiety symptomatology in young adults, PADu-multicenter (2021–2022)

Isotemporal models	SB	MPA	VPA
	OR (95%CI)	OR (95%CI)	OR (95%CI)
5 min/day
Replace SB	Dropped	1.02 (1.01–1.03)	1.04 (1.03–1.05)
Replace MPA	0.98 (0.97–0.99)	Dropped	1.02 (1.00-1.03)
Replace VPA	0.96 (0.95–0.97)	0.98 (0.97-1.00)	Dropped
10 min/day
Replace SB	Dropped	1.04 (1.03–1.06)	1.08 (1.06–1.11)
Replace MPA	0.96 (0.95–0.97)	Dropped	1.04 (1.01–1.07)
Replace VPA	0.92 (0.90–0.95)	0.96 (0.94–0.99)	Dropped
15 min/day
Replace SB	Dropped	1.06 (1.04–1.08)	1.12 (1.08–1.17)
Replace MPA	0.94 (0.92–0.96)	Dropped	1.06 (1.01–1.10)
Replace VPA	0.89 (0.86–0.92)	0.95 (0.91–0.99)	Dropped
20 min/day
Replace SB	Dropped	1.08 (1.06–1.11)	1.17 (1.11–1.23)
Replace MPA	0.92 (0.90–0.95)	Dropped	1.08 (1.02–1.14)
Replace VPA	0.85 (0.81–0.90)	0.93 (0.88–0.98)	Dropped
25 min/day
Replace SB	Dropped	1.11 (1.07–1.14)	1.22 (1.15–1.29)
Replace MPA	0.90 (0.87–0.93)	Dropped	1.10 (1.02–1.18)
Replace VPA	0.82 (0.77–0.87)	0.91 (0.85–0.98)	Dropped
30 min/day
Replace SB	Dropped	1.13 (1.09–1.18)	1.27 (1.18–1.36)
Replace MPA	0.88 (0.85–0.92)	Dropped	1.12 (1.03–1.22)
Replace VPA	0.79 (0.74–0.85)	0.89 (0.82–0.97)	Dropped
35 min/day
Replace SB	Dropped	1.15 (1.10–1.21)	1.32 (1.21–1.43)
Replace MPA	0.87 (0.83–0.91)	Dropped	1.14 (1.03–1.26)
Replace VPA	0.76 (0.70–0.83)	0.88 (0.79–0.97)	Dropped
40 min/day
Replace SB	Dropped	1.18 (1.12–1.24)	1.37 (1.24–1.51)
Replace MPA	0.85 (0.81–0.89)	Dropped	1.16 (1.04–1.30)
Replace VPA	0.73 (0.66–0.80)	0.86 (0.77–0.96)	Dropped
45 min/day
Replace SB	Dropped	1.20 (1.13–1.27)	1.42 (1.28–1.59)
Replace MPA	0.83 (0.79–0.88)	Dropped	1.18 (1.04–1.35)
Replace VPA	0.70 (0.63–0.78)	0.84 (0.74–0.96)	Dropped
50 min/day
Replace SB	Dropped	1.23 (1.15–1.31)	1.48 (1.31–1.67)
Replace MPA	0.82 (0.76–0.87)	Dropped	1.21 (1.05–1.39)
Replace VPA	0.68 (0.60–0.76)	0.83 (0.72–0.96)	Dropped
55 min/day
Replace SB	Dropped	1.25 (1.16–1.34)	1.54 (1.35–1.76)
Replace MPA	0.80 (0.74–0.86)	Dropped	1.23 (1.05–1.44)
Replace VPA	0.65 (0.57–0.74)	0.81 (0.69–0.95)	Dropped
60 min/day
Replace SB	Dropped	1.28 (1.18–1.38)	1.60 (1.39–1.85)
Replace MPA	0.78 (0.72–0.85)	Dropped	1.25 (1.06–1.49)
Replace VPA	0.62 (0.54–0.72)	0.80 (0.67–0.95)	Dropped

Note: CI = confidence interval; OR = odds ratio; MPA = moderate physical activity; VPA = vigorous physical activity; SB = sedentary behavior

All models showed p-values < 0.05 in the logistic regression model

Table 4.

Isotemporal substitution models of movement behaviors in anxiety symptomatology in young adults, PADu-multicenter (2021–2022)

Isotemporal models	SB	MPA	VPA
	OR (95%CI)	OR (95%CI)	OR (95%CI)
5 min/day
Replace SB	Dropped	1.03 (1.02–1.04)	1.04 (1.02–1.05)
Replace MPA	0.97 (0.96–0.98)	Dropped	1.01 (0.99–1.02)*
Replace VPA	0.97 (0.95–0.98)	0.99 (0.98–1.01)*	Dropped
10 min/day
Replace SB	Dropped	1.06 (1.05–1.08)	1.07 (1.05–1.10)
Replace MPA	0.94 (0.93–0.95)	Dropped	1.01 (0.98–1.04)*
Replace VPA	0.93 (0.91–0.95)	0.99 (0.96–1.02)*	Dropped
15 min/day
Replace SB	Dropped	1.09 (1.07–1.11)	1.11 (1.07–1.15)
Replace MPA	0.91 (0.90–0.93)	Dropped	1.02 (0.97–1.06)*
Replace VPA	0.90 (0.87–0.93)	0.98 (0.94–1.3)*	Dropped
20 min/day
Replace SB	Dropped	1.13 (1.10–1.16)	1.15 (1.10–1.21)
Replace MPA	0.89 (0.86–0.91)	Dropped	1.02 (0.97–1.08)*
Replace VPA	0.87 (0.83–0.91)	0.98 (0.92–1.04)*	Dropped
25 min/day
Replace SB	Dropped	1.16 (1.12–1.20)	1.19 (1.12–1.27)
Replace MPA	0.86 (0.83–0.89)	Dropped	1.03 (0.96–1.10)*
Replace VPA	0.84 (0.79–0.89)	0.97 (0.91–1.05)*	Dropped
30 min/day
Replace SB	Dropped	1.19 (1.15–1.24)	1.23 (1.15–1.33)
Replace MPA	0.84 (0.80–0.87)	Dropped	1.03 (0.95–1.12)*
Replace VPA	0.81 (0.75–0.87)	0.97 (0.89–1.05)*	Dropped
35 min/day
Replace SB	Dropped	1.23 (1.18–1.29)	1.28 (1.18–1.39)
Replace MPA	0.81 (0.78–0.85)	Dropped	1.04 (0.94–1.15)*
Replace VPA	0.78 (0.72–0.85)	0.96 (0.87–1.06)*	Dropped
40 min/day
Replace SB	Dropped	1.27 (1.20–1.34)	1.32 (1.20–1.46)
Replace MPA	0.79 (0.75–0.83)	Dropped	1.04 (0.93–1.17)*
Replace VPA	0.76 (0.69–0.83)	0.96 (0.85–1.07)*	Dropped
45 min/day
Replace SB	Dropped	1.31 (1.23–1.39)	1.37 (1.23–1.53)
Replace MPA	0.77 (0.72–0.81)	Dropped	1.05 (0.92–1.19)*
Replace VPA	0.73 (0.65–0.81)	0.95 (0.84–1.08)*	Dropped
50 min/day
Replace SB	Dropped	1.35 (1.26–1.44)	1.42 (1.26–1.60)
Replace MPA	0.74 (0.70–0.79)	Dropped	1.06 (0.92–1.22)*
Replace VPA	0.70 (0.62–0.79)	0.95 (0.82–1.09)*	Dropped
55 min/day
Replace SB	Dropped	1.39 (1.29–1.49)	1.47 (1.29–1.68)
Replace MPA	0.72 (0.67–0.78)	Dropped	1.06 (0.91–1.24)*
Replace VPA	0.68 (0.60–0.78)	0.94 (0.81–1.10)*	Dropped
60 min/day
Replace SB	Dropped	1.43 (1.32–1.55)	1.52 (1.32–1.76)
Replace MPA	0.70 (0.65–0.76)	Dropped	1.07 (0.90–1.27)*
Replace VPA	0.66 (0.57–0.76)	0.94 (0.79–1.11)*	Dropped

Note: CI = confidence interval; OR = odds ratio; MPA = moderate physical activity; VPA = vigorous physical activity; SB = sedentary behavior

* p-values > 0.05 in the logistic regression model

Discussion

The results of this study indicate a dose-response association between movement behaviors and symptoms of anxiety and depression in young adults in Brazil. Furthermore, our findings suggest that replacing time dedicated to SB with MPA and VPA is associated with positive effects on mental health, confirming our initial hypothesis.

The benefits of PA practice are widely recognized in the scientific literature, being directly associated with the promotion of mental health and the reduction of the risk of symptoms of anxiety and depression [18, 47–49]. Evidence indicates that physically active individuals have better mental health outcomes [49, 50]. In contrast, SB, has been identified as a risk factor for mental disorders in the population [47, 49, 51, 52]. This contrast between the effects of PA and SB justifies the growing interest in understanding the impact of movement behaviors on the mental health of young adults.

Despite the many health benefits of regular PA, the prevalence of SB and physical inactivity remains high, particularly during the college years. This period is marked by a significant reduction in PA and an increase in SB levels, which aggravates physical and mental health risks [18, 50, 53]. These concerning trends were further intensified during the COVID-19 pandemic, when public health measures imposed by governments—both at individual and institutional levels—drastically altered daily routines [54, 55]. Evidence suggests that the pandemic acted as a major stressor, significantly affecting the movement behaviors of university students and leading to substantial negative impacts on their physical and mental well-being [56, 57].

Several studies have shown that, during this period, university students experienced a reduction in PA levels along with a marked increase in SB, especially when compared to the general population [56, 58]. Supporting this, a longitudinal study reported significant reductions in moderate-to-vigorous physical activity (MVPA) levels and increased SB among university students during the COVID-19 restrictions compared to the pre-pandemic period [58]. Similarly, Gallè et al. [59] observed a substantial rise in time spent in SB, particularly involving electronic devices, and a decrease in total PA among university students during home confinement.

These findings are consistent with the results of the present study, which also identified a high prevalence of SB (55.09%) among university students. It is likely that the academic environment significantly contributes to this scenario, as it involves various activities that require prolonged sitting, such as attending classes and engaging in study-related tasks. Moreover, remote learning may have further intensified SB by reducing daily opportunities for movement [20, 21].

Evidence suggests that replacing SB with PA and VPA is associated with a range of health outcomes, including a reduced risk of morbidity and mortality, indicators of body adiposity including BMI, cardiometabolic biomarkers, and mental health among youths, adults, and older adults [14, 15, 24]. This reinforces the importance of reducing the time spent on SB even at low levels, and highlights the role of PA and VPA in health, as recommended by the WHO.

In a study conducted with overweight or obese Chinese university students, Wang et al. [3] observed that reallocating 15 min/day of light PA (LPA), SB in front of screens (such as watching TV and playing computer or video games), and SB away from screens (e.g., sitting and reading) to MVPA and sleep were significantly associated with lower scores for symptoms of depression, anxiety, and stress, as assessed by the DASS-21. These findings are consistent with the results of the present study, as well as with those of a systematic review and meta-analysis that aimed to synthesize evidence from 17 studies using isotemporal substitution models to estimate the potential effects of reallocating time spent in SB to different intensities of PA in adults with depression. The authors found that replacing sedentary time with MVPA and sleep was associated with a significant reduction in depressive symptoms [60].

Several mechanisms can explain the influence of PA and SB on the development of mental disorders. Among the various effects of regular PA on mental health, increased neurogenesis in the hippocampus is the neurochemical phenomenon most strongly associated with the impact of exercise on the CNS. Exercise increases the number of new neurons and influences the morphology of these newly formed neurons [61].

Concomitant with neurogenesis, activation of the HPA, a neuroendocrine system responsible for regulating stress hormones such as cortisol, occurs in the body [18]. When stimulated, the hypothalamus secretes corticotropin-releasing hormone, which stimulates the pituitary gland to synthesize adrenocorticotropic hormone (ACTH). This process increases glucocorticoids (cortisol) as ACTH interacts with the adrenal gland [62]. The literature indicates that individuals with mental disorders have dysregulation of the HPA axis [63]. In this context, regular PA practice can act as a moderator, helping to maintain proper functioning of the HPA axis, reducing cortisol levels, and protecting against harmful effects on mental health, especially among sedentary individuals [18, 63].

The mechanisms presented above regarding PA also apply to SB, which, by its nature, constitutes a significant risk factor for the development of mental disorders. Among the underlying mechanisms that explain the relationship between SB and mental health, the systemic inflammatory process stands out, which can be attributed, for example, to the accumulation of abdominal fat in visceral adipose tissue [64]. In the chronic inflammation typical of obesity, high levels of cortisol can affect the negative feedback mechanism responsible for regulating the activity of the HPA axis, which could potentially increase levels of anxiety and depression [64, 65].

In addition to the effect of cortisol, evidence suggests that the association between SB, especially screen-based activities, and mental health can also be explained by direct cognitive effects, such as low emotional stability and impulsivity. Furthermore, SB can interfere with interpersonal relationships, leading to social isolation and feelings of loneliness, which negatively impact mental health, favoring the development of mental disorders [19]. These mechanisms highlight the importance of reducing SB levels and encouraging PA practice to promote significant mental health benefits.

The limitations of this study should be considered when interpreting the results. First, the cross-sectional design did not allow us to infer cause-and-effect relationships between movement behaviors and mental health. Additionally, subjective measures were used to assess the levels of PA and SB, which may have introduced recall biases, resulting in the overestimation or underestimation of data. Notably, the scale used to assess symptoms of anxiety and depression, the DASS-21, is a self-reported instrument that does not provide a medical diagnosis but rather a score that classifies the levels of the symptoms assessed. Additionally, data collection occurred during the COVID-19 pandemic, a period that may have influenced both movement behaviors and perceptions of participants’ feelings. For example, the transition to remote learning has impacted the routines and lifestyles of the university population and mental health. Finally, our study did not incorporate all daily movement-related behaviors, such as sleep, in the isotemporal analysis due to the absence of a specific question about sleep duration in the questionnaire used. Future studies should aim to include all three behaviors (sleep, PA and SB) in the model to enhance the understanding of their interactions and their impact on mental health. Longitudinal studies are recommended to explore a wider range of health outcomes associated with the isotemporal substitutions of SB, PA, and sleep.

The present study had several strengths. To the best of our knowledge, this is the first study conducted in Brazil to incorporate statistical isotemporal substitution modeling to assess the hypothetical effects of reallocating a given amount of time spent in one movement behavior to another and its impact on mental health. This approach may improve the understanding of the interrelationships between different movement behaviors and their health consequences and contribute to formulating guidelines and strategies for public health promotion. Another strength of our study is the sample size and multicenter approach adopted. We conducted this study using a robust sample of eight Brazilian public universities from different areas of knowledge. This approach is relatively uncommon in studies involving university students. However, it is important to note that all participating institutions are located in a single Brazilian state. Given the socioeconomic, cultural, and educational diversity across the 27 Brazilian states, this geographic concentration may limit the generalizability of our findings to university students from other regions of the country. An essential difference in this study is that, unlike most studies that focus only on students in the health field, such as medical students, we included students from three areas of knowledge: life sciences, exact sciences, humanities, social sciences, and applied sciences. These characteristics strengthen the reliability of the results and broaden their representativeness. Thus, our study contributes to a more comprehensive understanding of movement behaviors and mental health among university students.

Conclusion

The findings of this study suggest that the assessed movement behaviors are associated with the presence symptoms of anxiety and depressive in young adults. Furthermore, replacing time spent on SB with an equivalent amount of MPA or VPA may reduce the likelihood of developing mental disorders. Thus, the greater the reallocation of time spent on movement behaviors, the greater the observed benefits. The approach adopted in this study expands the evidence on movement behaviors over 24 h and their association with mental health. Increasing PA levels and reducing SB should be a public health priority to promote mental health and improve the physical health of students who present with symptoms of mental disorders. In this sense, reducing SB whenever possible is encouraged, and taking regular breaks during prolonged periods of sitting is recommended as a health promotion strategy. For example, standing up, stretching, walking to get water, or using the restroom for at least five minutes every hour can help interrupt prolonged sedentary time and improve overall well-being. In addition, individuals who spend a large portion of the day sitting, such as university students, can be encouraged to compensate for this time by incorporating more physical activity into their daily routines.

Author contributions

BCRB– coordination and conducted of data collection, analysis and interpretation of data, writing of the manuscript and critical review of the manuscript. LAAMJ– contributed to analysis and interpretation of data and critical review of the manuscript. LFF and MCAV– writing of the manuscript and critical review of the manuscript. WP, CMSC, ELM, LNN, EDF, FCV, CSC and LSS– coordination and conducted of data collection and critical review of the manuscript. ALM– conception and coordination of data collection, conception and study design and critical review of the manuscript. All authors have read and approved the final version of the manuscript, and agree with the order of presentation of the authors.

Funding

The research project “Symptoms of anxiety disorder and depression among university students in Minas Gerais: prevalence and associated factors” was supported by Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG/Brazil) under PROCESS number N. CDS - APQ-01089- 18, with granting term FAPEMIG/DOF nº. 8288757/2019. This study was supported by the FAPEMIG/Brazil with PhD student scholarship.

Data availability

The datasets generated and/or analyzed as part of the current study are not publicly available due to confidentiality agreements with subjects. However, they can be made available solely for the purpose of review and not for the purpose of publication from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Research Ethics Committee of the Federal University of Ouro Preto, under protocol number 31077320.7.1001.5150, and by the Ethics Committees of all participating universities (Federal University of Minas Gerais 43027421.3.2004.5149; Federal University of Uberlândia: 43027421.3.2001.5152; Juiz de Fora Federal University: 43027421.3.2003.5147; Federal University of São João del-Rei: 43027421.3.2002.5545; Federal University of Lavras: 43027421.3.2006.5148; Federal University of the Jequitinhonha and Mucuri Valleys: 43027421.3.2009.5108; Federal University of Alfenas: 43027421.3.2008.5142). All procedures adopted in this study followed the Declaration of Helsinki and the Brazilian guidelines and norms for research involving humans. Informed consent was obtained from all individual participants included for study participation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.