Changes in Aerobic and Muscle-Strengthening Physical Activity and Self-Reported Morbidity Among University Students During the COVID-19 Pandemic: A Cross-Sectional Study

Highlights

Public health relevance—How does this work relate to a public health issue?

• University students represent an important population for health promotion, as health-related behaviors developed during this life stage may persist into adulthood.
• The COVID-19 pandemic disrupted university students’ routines and physical activity behaviors, creating a relevant context for study associations between changes in physical activity and cardiometabolic morbidity.

Public health significance—Why is this work of significance to public health?

• The study separately examined changes in moderate-intensity aerobic physical activity (MPA) and muscle-strengthening physical activity (MSPA), allowing modality-specific associations with self-reported morbidity to be investigated.
• In separate adjusted models, both MPA and MSPA were associated with morbidity; however, these associations were attenuated when both activity modalities were included in the same model.

Public health implications—What are the key implications or messages for practitioners, policy makers and/or researchers in public health?

• University health research and surveillance may benefit from considering aerobic and muscle-strengthening activities separately rather than treating physical activity as a single behavior.
• Longitudinal studies are needed to clarify the temporal direction of the observed associations and the independent contributions of aerobic and muscle-strengthening physical activity to cardiometabolic health.

Abstract

The COVID-19 pandemic substantially affected lifestyle behaviors among university students, potentially increasing their vulnerability to chronic noncommunicable diseases (NCDs). This cross-sectional study investigated associations between self-reported changes in the frequency of moderate-intensity aerobic physical activity (MPA) and muscle-strengthening physical activity (MSPA) and self-reported morbidity among university students. Data were collected through an online questionnaire from 1150 students at a private university in Rio de Janeiro, Brazil, between August 2021 and April 2022. Morbidity was defined as a self-reported medical diagnosis of hypertension, type 2 diabetes mellitus, and/or cardiovascular disease. Separate and mutually adjusted binary logistic regression models were controlled for sex, age, smoking, sugar consumption, and fat consumption. Morbidity prevalence was 9.4%. In separate adjusted models, MPA (global p = 0.046) and MSPA (global p = 0.016) were associated with morbidity. Compared with no engagement, all MPA engagement categories were linked to lower odds of morbidity (ORs = 0.41–0.42), as were increased (OR = 0.41; 95% CI: 0.22–0.75) and maintained MSPA (OR = 0.44; 95% CI: 0.24–0.81). In the mutually adjusted model, the overall associations were attenuated for MPA (p = 0.442) and MSPA (p = 0.127), although increased MSPA remained individually associated with lower odds of morbidity (OR = 0.42; 95% CI: 0.19–0.93; p = 0.032). These findings should be interpreted cautiously, as the cross-sectional design precludes establishing temporal direction and reverse causality is possible.

1. Introduction

Chronic noncommunicable diseases (NCDs) represent one of the major challenges to global public health, as they are associated with a substantial burden of morbidity, mortality [1] and socioeconomic costs [2].

Although NCDs are more common among middle-aged and older adults, recent evidence indicates that cardiometabolic risk factors and some NCDs are already present in young adults, reflecting changes in lifestyle-related behavioral patterns [3,4,5]. In this context, understanding the factors associated with the early occurrence of these conditions has become a public health priority. Studies in university populations have also reported hypertension, diabetes mellitus, and other cardiometabolic alterations, indicating that these conditions are already present during this stage of adulthood [6,7,8].

The transition to higher education is characterized by profound changes in daily routines. Greater autonomy, academic demands, and changes in dietary habits, sleep, and physical activity promote the adoption of behaviors that may influence both current and future health [9,10]. Since many of these behaviors tend to persist throughout adulthood, university students represent a strategic population for research and the implementation of health promotion initiatives. Moreover, university populations include students across a broad age range and at different stages of adulthood, allowing for the investigation of health-related behaviors and morbidity within a heterogeneous higher education setting.

Among the modifiable risk factors related to the prevention of NCDs, physical activity stands out as one of the most critical. Consistent evidence indicates that physically active individuals have a lower risk of hypertension, type 2 diabetes mellitus, cardiovascular disease, and all-cause mortality [1,11,12]. However, physical activity is not a single, homogeneous behavior. International guidelines emphasize that aerobic physical activity and muscle-strengthening exercises provide complementary benefits and operate through partially distinct physiological mechanisms, which is why both should be incorporated into a healthy lifestyle [1]. Despite this, epidemiological studies frequently examine physical activity as a single construct, making it difficult to understand the specific contribution of different modalities to morbidity.

This scenario became particularly relevant during the COVID-19 pandemic. Social distancing measures and the transition to remote learning profoundly altered the daily routines of university students, leading to important changes in physical activity and other lifestyle-related behaviors [13,14,15,16,17]. A previous study conducted in the same university population during the pandemic also identified associations between changes in physical activity and self-rated health [18]. Although these changes have been widely documented, evidence regarding their association with the presence of morbidity remains limited, particularly when self-reported changes in the frequency of moderate-intensity aerobic and muscle-strengthening physical activity are considered separately. Examining these modalities separately may therefore help clarify whether their associations with morbidity differ within the university population, an aspect that remains insufficiently explored in the pandemic context.

To address this gap in the literature, the present study aimed to examine whether self-reported changes in the frequency of moderate-intensity aerobic physical activity (MPA) and muscle-strengthening physical activity (MSPA) were associated with self-reported morbidity among university students during the COVID-19 pandemic. We hypothesized that (1) students who maintained or increased MPA would have lower odds of self-reported morbidity than those who did not engage in MPA, and (2) students who maintained or increased MSPA would have lower odds of self-reported morbidity than those who did not engage in MSPA.

2. Materials and Methods

2.1. Study Design and Population

This was an observational cross-sectional study conducted among university students enrolled at a private higher education institution in the state of Rio de Janeiro, Brazil (Estácio de Sá University). Data were collected between August 2021 and April 2022, during which classes were delivered remotely to comply with social distancing measures implemented in response to the COVID-19 pandemic. Information was obtained using an electronic questionnaire, ensuring participants’ anonymity and confidentiality. All participants provided written informed consent. The study was approved by the Research Ethics Committee of Estácio de Sá University (UNESA) under approval number 4.844.578.

The target population comprised undergraduate students enrolled in on-campus programs who were aged 18 years or older and were actively enrolled during the second semester of 2021 and the first semester of 2022. According to institutional records from May 2021, the target population was estimated at 110,023 students. Individuals younger than 18 years, those with inactive enrollment, and students enrolled in distance-learning programs were excluded.

The sample size was calculated using the formula for finite populations [19], assuming a 95% confidence level, a 3-percentage-point margin of error, and a prevalence of 50%, resulting in a minimum required sample of 1057 participants. A prevalence of 50% was assumed because it provides the most conservative estimate of the required sample size when the expected prevalence is unknown. This calculation was intended to ensure precision in estimating the prevalence in the target population and was not an a priori power calculation for the multivariable regression analyses. The final study sample comprised 1150 university students. The completed STROBE checklist is provided in the Supplementary Materials.

2.2. Procedures

An electronic questionnaire developed on Google Forms was used for data collection. The survey link was sent by the research team to undergraduate program coordinators, who were asked to forward it to students who were regularly enrolled in on-campus undergraduate programs, although classes were being delivered remotely during the study period. Each coordinator determined the method of distribution, which could include WhatsApp and/or email. Because the invitations were distributed in a decentralized manner and the research team did not have access to the number of students who received or viewed the invitation, the number of individuals invited and the response rate could not be determined.

The data collection instrument included a survey comprising sociodemographic, academic, and health-related questions, as well as the previously validated PERMEV questionnaire (Percepção das mudanças no estilo de vida durante o distanciamento social [Perception of Changes in Lifestyle during Social Distancing]). The original PERMEV comprised 40 items organized into eight domains. Following its validation, six items were excluded according to the exploratory factor analysis criteria, resulting in a final structure of 34 items distributed across 10 factors [20]. The present study used this validated 34-item structure.

The outcome was the presence of self-reported morbidity, defined as a self-reported physician diagnosis of hypertension, type 2 diabetes mellitus, and/or cardiovascular disease, ascertained using the following question: “Has a doctor ever diagnosed you with any of these conditions?” Participants could select more than one response option. A dichotomous variable was subsequently created to indicate the absence (0) or presence (1) of at least one of the assessed conditions.

These conditions were grouped, as they constitute major chronic noncommunicable diseases and share modifiable risk factors [1]. The composite outcome was used because the individual conditions were relatively infrequent in this university population and the study aimed to examine the presence of cardiometabolic morbidity rather than condition-specific associations.

The exposures were moderate-intensity aerobic physical activity and muscle-strengthening physical activity, assessed using two items from the PERMEV questionnaire regarding changes in the frequency of these activities during the COVID-19 pandemic. The questions were: (i) “Compared with the initial period of the COVID-19 pandemic, how often have you engaged in leisure-time moderate physical activity (such as walking, running, cycling, dancing, or similar activities that make you breathe somewhat harder than normal) during the past seven days?”; and (ii) “Compared with the initial period of the COVID-19 pandemic, how often have you engaged in muscle-strengthening physical activity (using your own body weight, weight plates, dumbbells, resistance bands, or household items like plastic bottles and food bags) during the past seven days?” Response options were: “much less often than before”, “less often than before”, “about the same”, “more often than before”, “much more often than before”, and “I do not engage in this activity”. For statistical analyses, responses were categorized as: Do not engage; Maintained; Increased (“more often” and “much more often”); and Decreased (“less often” and “much less often”).

Changes in sugar and fat consumption were also assessed using specific items from the PERMEV questionnaire. Participants reported changes in the frequency of consumption of foods high in sugar and foods high in fat during the COVID-19 pandemic, using the same response categories applied to the physical activity items. For the analyses, responses were categorized as: Do not consume; Maintained; Increased; and Decreased.

For the multivariable analyses, sex, age, smoking, sugar consumption, and fat consumption were included as covariates based on prior evidence [21] and their theoretical plausibility as potential confounding factors [22,23]. Age was entered as a continuous variable to minimize residual confounding.

2.3. Data Analysis

Statistical analyses were performed using SPSS software, version 23. Initially, descriptive analyses were conducted by calculating the absolute and relative frequencies of the categorical variables.

Next, associations were estimated using binary logistic regression, with odds ratios (ORs) and their corresponding 95% confidence intervals (95% CIs) calculated for the adjusted models. The “Do not engage” category was used as the reference category for both MPA and MSPA.

To reduce model complexity and examine the associations of the two physical activity modalities, three multivariable models were estimated: one model including MPA and the selected covariates, one including MSPA and the selected covariates, and a mutually adjusted model including both MPA and MSPA together with the same covariates.

The association between MPA and MSPA was examined using Cramer’s V, and multicollinearity was assessed using variance inflation factors (VIFs). Cramer’s V was 0.587 (p < 0.001), indicating substantial overlap between the two physical activity measures, while VIF values ranged from 1.027 to 5.220, indicating moderate but not severe multicollinearity.

Model calibration was assessed using the Hosmer–Lemeshow goodness-of-fit test, and explained variation was summarized using Nagelkerke’s R². Discrimination of each logistic regression model was evaluated using the area under the receiver operating characteristic curve (AUC).

Analyses were conducted using complete cases for the variables included in each model; no missing-data imputation was performed because the proportion of missing observations was low (1.5% in the MPA and mutually adjusted models and 1.3% in the MSPA model). The corresponding analytical samples were 1133, 1135, and 1133 participants, respectively. A significance level of 5% (p < 0.05) was adopted.

3. Results

The sample comprised 1150 university students, predominantly female (63.8%), aged 25 years or older (61.5%), residing in the state capital (84.4%), enrolled in Health Sciences programs (52.3%), enrolled in four or more courses (84.4%), and attending evening classes (66.3%).

Regarding health behaviors, most students reported not smoking (79.1%) and maintaining their consumption of foods high in fat (45.8%) and sugar (44.2%) when compared with the initial period of the pandemic.

As for physical activity, 35.5% of participants reported an increase in MPA, whereas 30.5% reported an increase in MSPA during the pandemic. Self-reported morbidities were identified in 108 students (9.4%).

The sociodemographic, academic, and behavioral characteristics of the study participants, including the exposures and the outcome, are presented in

Table 1. Sociodemographic, academic, and behavioral characteristics of study participants (n = 1150).

Variable	n	%
Biological sex (n = 1150)
Female	734	63.8
Male	416	36.2
Age (years), mean ± SD: 30.73 ± 10.79 (n = 1146)
18–24 years	441	38.5
≥25 years	705	61.5
Place of residence (n = 1150)
State capital	971	84.4
Outside the state capital	179	15.6
Field of study (n = 1140)
Health sciences	596	52.3
Other fields	544	47.7
Number of enrolled courses (n = 1141)
≤3	178	15.6
≥4	963	84.4
Class schedule (n = 1146)
Daytime	386	33.7
Evening	760	66.3
Smoking (n = 1150)
Decreased	68	5.9
Increased	43	3.7
Maintained	129	11.2
Non-smoker	910	79.1
Consumption of high-fat foods (n = 1147)
Decreased	251	21.9
Increased	335	29.2
Maintained	525	45.8
Do not consume	36	3.1
Consumption of high-sugar foods (n = 1146)
Decreased	284	24.8
Increased	314	27.4
Maintained	507	44.2
Do not consume	41	3.6
MPA (n = 1147)
Decreased	295	25.7
Increased	407	35.5
Maintained	341	29.7
Do not engage	104	9.1
MSPA (n = 1150)
Decreased	257	22.3
Increased	351	30.5
Maintained	315	27.4
Do not engage	227	19.7
Morbidity (n = 1150)
Yes	108	9.4
No	1042	90.6

Note: Denominators vary because of missing data. MPA = moderate-intensity aerobic physical activity; MSPA = muscle-strengthening physical activity.

.

Among participants with valid data for both the physical activity exposure and morbidity outcome, the distribution of morbidity across the MPA categories was 27/295 (9.2%) among students who decreased MPA, 33/406 (8.1%) among those who increased MPA, 30/341 (8.8%) among those who maintained MPA, and 18/104 (17.3%) among those who did not engage in MPA. For MSPA, morbidity was present in 26/257 (10.1%) students who decreased MSPA, 23/350 (6.6%), who increased MSPA, 26/315 (8.3%) who maintained MSPA, and 33/227 (14.5%) who did not engage in MSPA.

In the crude analyses, lower odds of morbidity were observed for all MPA categories compared with the “Do not engage” category, whereas for MSPA, lower odds were observed for the increased and maintained categories (

Table 2. Associations of changes in MPA and MSPA with self-reported morbidity in separate and mutually adjusted logistic regression models.

Exposure	Category	Crude OR (95% CI)	Separate Model OR (95% CI)	p	Mutually Adjusted Model OR (95% CI)	p
MPA	Decreased	0.48 (0.25–0.92)	0.416 (0.206–0.841)	0.015	0.497 (0.200–1.235)	0.132
	Increased	0.42 (0.23–0.79)	0.423 (0.213–0.840)	0.014	0.789 (0.327–1.905)	0.598
	Maintained	0.46 (0.24–0.87)	0.405 (0.202–0.811)	0.011	0.666 (0.277–1.599)	0.363
	Do not engage	1.00	1.00		1.00
	Global test			0.046		0.442
MSPA	Decreased	0.66 (0.38–1.14)	0.581 (0.320–1.055)	0.074	0.845 (0.391–1.826)	0.668
	Increased	0.41 (0.24–0.72)	0.405 (0.218–0.752)	0.004	0.421 (0.191–0.927)	0.032
	Maintained	0.53 (0.31–0.91)	0.442 (0.241–0.809)	0.008	0.515 (0.241–1.100)	0.086
	Do not engage	1.00	1.00		1.00
	Global test			0.016		0.127

Note: MPA = moderate-intensity aerobic physical activity; MSPA = muscle-strengthening physical activity; OR = odds ratio; CI = confidence interval. “Do not engage” was the reference category. Separate models included either MPA or MSPA and were adjusted for sex, age (continuous), smoking, sugar consumption, and fat consumption. The mutually adjusted model included MPA and MSPA simultaneously and was adjusted for the same covariates. Complete-case analysis: MPA separate model, n = 1133; MSPA separate model, n = 1135; mutually adjusted model, n = 1133.

).

The results of the separate and mutually adjusted logistic regression models are presented in Table 2. In the separate adjusted model for MPA, the overall association with morbidity was statistically significant (global p = 0.046). Compared with students who did not engage in MPA, lower odds of morbidity were observed among those who reported decreased (OR = 0.416; 95% CI: 0.206–0.841; p = 0.015), increased (OR = 0.423; 95% CI: 0.213–0.840; p = 0.014), or maintained MPA (OR = 0.405; 95% CI: 0.202–0.811; p = 0.011). The similarity of these estimates suggests no clear gradient according to the direction of change in MPA.

In the separate adjusted model for MSPA, the overall association with morbidity was also statistically significant (global p = 0.016). Compared with students who did not engage in MSPA, increased (OR = 0.405; 95% CI: 0.218–0.752; p = 0.004) and maintained MSPA (OR = 0.442; 95% CI: 0.241–0.809; p = 0.008) were associated with lower odds of morbidity, whereas decreased MSPA was not statistically significant (OR = 0.581; 95% CI: 0.320–1.055; p = 0.074).

When MPA and MSPA were included simultaneously, the overall associations were attenuated and were no longer statistically significant for MPA (global p = 0.442) or MSPA (global p = 0.127). However, increased MSPA remained associated with lower odds of morbidity (OR = 0.421; 95% CI: 0.191–0.927; p = 0.032). No MPA category remained individually associated with morbidity in the mutually adjusted model.

The separate MPA model showed adequate calibration (Hosmer–Lemeshow p = 0.577; Nagelkerke R² = 0.173) and acceptable discrimination (AUC = 0.771; 95% CI: 0.724–0.818; p < 0.001), as did the separate MSPA model (Hosmer–Lemeshow p = 0.721; Nagelkerke R² = 0.178; AUC = 0.768; 95% CI: 0.719–0.816; p < 0.001). The mutually adjusted model also showed adequate calibration (Hosmer–Lemeshow p = 0.721; Nagelkerke R² = 0.183) and acceptable discrimination (AUC = 0.772; 95% CI: 0.723–0.820; p < 0.001).

4. Discussion

The present study investigated the associations between self-reported changes in the frequency of MPA and MSPA and self-reported morbidity among university students during the COVID-19 pandemic. In separate adjusted models, both MPA and MSPA were associated with morbidity. For MPA, all categories of engagement showed similarly lower odds of morbidity compared with no engagement, with no clear gradient according to whether activity was decreased, increased, or maintained. For MSPA, increased and maintained activity were associated with lower odds of morbidity compared with no engagement. When MPA and MSPA were included simultaneously, however, the overall associations for both modalities were attenuated and were no longer statistically significant, although increased MSPA remained individually associated with lower odds of morbidity. These findings indicate that the associations of MPA and MSPA with morbidity are partly overlapping and should be interpreted cautiously, particularly given the cross-sectional design and the possibility of reverse causality.

The prevalence of self-reported morbidity in the present study was 9.4%. Although relatively low, this finding indicates that hypertension, type 2 diabetes mellitus, and cardiovascular disease were present among university students in the study population. The relatively low prevalence may partly reflect the characteristics of the sample and the age distribution of university populations; nevertheless, the present sample was heterogeneous in age, with 61.5% of participants aged 25 years or older. Therefore, the findings should not be interpreted as representing exclusively young adults or the traditional university-age population.

Previous studies conducted in different countries have demonstrated the occurrence of hypertension, diabetes mellitus, and other cardiometabolic alterations among university populations, supporting the relevance of cardiometabolic health in this population [6,7,8]. Although methodological differences and differences in population characteristics limit direct comparisons between studies, these findings support the importance of considering NCDs in health research and health promotion initiatives targeting university populations.

An important finding was the pattern observed for MPA in the separate adjusted model. Compared with students who did not engage in MPA, those who reported decreased, increased, or maintained MPA showed similarly lower odds of self-reported morbidity, with OR estimates ranging from 0.405 to 0.423. The similarity of these estimates does not support a dose–response pattern according to the direction of change in MPA. Instead, the findings suggest that the observed association may primarily distinguish students who engaged in MPA from those who did not engage in this activity, irrespective of whether its frequency decreased, increased, or remained stable during the pandemic. Nonetheless, because the study assessed changes in activity retrospectively and morbidity may have preceded these changes, this pattern should not be interpreted as evidence that engagement in MPA reduced the occurrence of morbidity.

When MPA and MSPA were included simultaneously in the model, their overall associations with morbidity were attenuated and were no longer statistically significant. MPA and MSPA were substantially associated with each other (Cramer’s V = 0.587), and the VIF values indicated moderate collinearity between some of their categories. Thus, the attenuation observed in the mutually adjusted model may partly reflect shared information between the two physical activity modalities, making their independent associations with morbidity more difficult to distinguish [24]. These findings do not establish that collinearity alone explains the attenuation, and other behavioral and health-related factors may also contribute to the observed associations. These findings therefore highlight the importance of considering both the separate and mutually adjusted associations of aerobic and muscle-strengthening physical activity when examining their relationships with health outcomes.

Although the overall association for MSPA was attenuated in the mutually adjusted model, increased MSPA remained individually associated with lower odds of self-reported morbidity. However, because the global test for MSPA was not statistically significant, this category-specific finding should be interpreted cautiously and does not establish an independent association between MSPA and morbidity. Recent evidence indicates that resistance training promotes physiological adaptations that extend beyond increases in muscular strength, encompassing improvements in insulin sensitivity, body composition, endothelial function, and blood pressure regulation, all of which are relevant to cardiovascular and metabolic health [25,26]. Although the mechanisms underlying this association cannot be confirmed in the present study, these physiological adaptations provide a plausible context for the observed association but should not be interpreted as mechanisms demonstrated by the present findings.

Another relevant aspect is that physical activity can be assessed as an overall construct or according to specific modalities. By separately analyzing moderate-intensity aerobic physical activity and muscle-strengthening physical activity, the present study allowed modality-specific patterns of association with self-reported morbidity to be examined. The differences observed between the separate and mutually adjusted models also indicate that considering these modalities separately may provide information that would be obscured if physical activity were treated as a single construct. The separate consideration of these activity modalities is also consistent with international recommendations that recognize muscle-strengthening activity as an essential component of physical activity for health promotion, rather than merely a complement to aerobic activities [1].

The observed associations are also consistent with previous evidence derived from the same study sample during the COVID-19 pandemic, using the same MPA and MSPA exposures, in which muscle-strengthening physical activity was associated with lower odds of negative self-rated health [18]. Because this previous analysis was based on the same sample and exposures, it should not be interpreted as an independent replication of the present findings. Nevertheless, although self-rated health and the presence of morbidity represent distinct health outcomes, the previous finding further supports the relevance of examining muscle-strengthening physical activity separately when investigating health-related outcomes among university students.

These findings should be interpreted in light of several limitations. First, the cross-sectional design does not allow the temporal sequence between changes in physical activity and morbidity to be established. The assessed morbidities were based on previous physician diagnoses and may therefore have preceded the changes in physical activity reported during the pandemic. Consequently, reverse causality is possible, as the presence of chronic condition, medical advice, or other health-related circumstances may have influenced participants’ physical activity behavior. The observed associations should therefore not be interpreted as evidence that changes in MPA or MSPA reduced or increased the occurrence of morbidity. In addition, the use of a composite morbidity outcome increased the number of events available for analysis but precludes condition-specific inferences regarding hypertension, type 2 diabetes mellitus, or cardiovascular disease.

Second, both morbidity and changes in physical activity were self-reported and may therefore be subject to recall and reporting bias. Although morbidity was based on self-reported physician diagnoses, no clinical measurements or medical records were available to confirm these conditions. Physical activity was not assessed using objective measures, and the instrument captured perceived changes in frequency rather than absolute levels, duration, or objectively measured intensity of MPA and MSPA. Body mass index and other indicators of adiposity were not available and could not be included as potential confounders. In addition, given the approximately 108 morbidity events and the number of parameters estimated in the adjusted models, the number of events per estimated parameter was below the conventional threshold of 10, which may have limited the precision and stability of the adjusted estimates, as reflected in some of the relatively wide confidence intervals. Furthermore, because recruitment was conducted through undergraduate program coordinators and the number of students who received or viewed the invitation was unknown, a response rate could not be calculated. This decentralized recruitment strategy may have introduced self-selection bias, and the direction and magnitude of this bias cannot be determined because information on nonrespondents was unavailable. The predominance of students from Health Sciences programs and the broad age distribution of the sample may also limit the generalizability of the findings to other university populations. Finally, the data were collected between 2021 and 2022, during a specific period of the COVID-19 pandemic characterized by remote learning and substantial disruptions to daily routines; therefore, the observed behavioral patterns and associations may not reflect those of university students in the current post-pandemic context.

Notwithstanding these limitations, key strengths of this study warrant emphasis. The analysis was conducted on a large sample of university students during a period of substantial behavioral changes amid the COVID-19 pandemic. In addition, the separate assessment of changes in MPA and MSPA represents a methodological strength, allowing the identification of modality-specific patterns of association that might have been obscured if only an overall measure of physical activity had been used. These results contribute to a better understanding of the relationships between different physical activity modalities and health among university students and provide a basis for future longitudinal investigations designed to clarify temporal relationships and the independent contributions of MPA and MSPA.

5. Conclusions

This study identified different patterns of association between self-reported changes in MPA and MSPA and self-reported morbidity among university students during the COVID-19 pandemic. In separate adjusted models, both MPA and MSPA were associated with morbidity. For MPA, similar lower odds of morbidity across all engagement categories compared with no engagement suggested no clear gradient according to the direction of change. For MSPA, increased and maintained activity were associated with lower odds of morbidity in the separate model. When both modalities were included simultaneously, their overall associations were attenuated and were no longer statistically significant, although increased MSPA remained individually associated with lower odds of morbidity.

These findings highlight the value of examining aerobic and muscle-strengthening physical activity separately while also considering their overlap when investigating associations with health outcomes. Given the cross-sectional design, the results should not be interpreted as evidence of a protective effect of physical activity or of a causal relationship between changes in physical activity and morbidity.

Longitudinal studies are needed to establish temporal relationships and to clarify the independent contributions of aerobic and muscle-strengthening physical activity to health outcomes among university students.