This study was approved by the Ethics Committee of Zunyi Medical University (Ethical Approval No.: 2025-1-111) and strictly adhered to the ethical principles outlined in the Declaration of Helsinki (14). The study employed a cross-sectional design, and all participants were fully informed about the study’s objectives, procedures, and potential risks before the study commenced, and provided written informed consent.

Inclusion criteria: (1) Adults aged 45 years and older; capable of understanding and agreeing to participate in the study and signing the informed consent form. (2) Residing in a relatively stable living environment for at least 1 month (to avoid the impact of short-term environmental changes on sleep and physical activity). (3) Capable of independently completing assessment tools such as the PSQI, GAD-7, and IPAQ; no severe physical illnesses or motor function disabilities, and able to perform daily physical activities. (4) No systemic psychological treatment or use of psychotropic medications affecting anxiety or sleep in the past 3 months. Exclusion criteria: (1) Diagnosed with severe mental disorders (e.g., schizophrenia, bipolar disorder). (2) Severe sleep disorders such as insomnia or sleep apnea syndrome (diagnosed). (3) Serious chronic diseases affecting physical activity (e.g., cardiovascular diseases, severe joint diseases). (4) Recent major life events (e.g., bereavement, significant financial crisis) that may impact anxiety levels, pregnant or breastfeeding women. (5) Recent (within the past 3 months) history of drug or alcohol abuse. (6) Unwilling or unable to cooperate with completing relevant questionnaires during the study.

The Pittsburgh PSQI was used to assess the sleep quality of participants in this study. The PSQI, developed by Buysse et al. (15), consists of 19 self-reported items and covers seven dimensions: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction (16). The score for each dimension ranges from 0 to 3, and the total score ranges from 0 to 21, with higher scores indicating poor sleep quality.

In addition, to further analyze sleep patterns, participants were asked to respond to a self-assessment question about sleep duration: “How many hours do you typically sleep on workdays or weeknights?” Based on national sleep duration recommendations, participants were categorized into three groups based on their sleep duration: short sleep (<7 h), normal sleep (7–9 h), and long sleep (≥9 h) (17). This classification method helps to further explore the relationship between different sleep durations, anxiety levels, and physical activity levels.

The short form of the IPAQ was used to assess participants’ physical activity levels. Participants were asked to recall and report the frequency and duration of high-intensity physical activity (HPA), moderate-intensity physical activity (MPA), walking (W), and sedentary behavior during the past 7 days. The physical activity levels were calculated in terms of Metabolic Equivalent of Task (MET), which estimates the energy expenditure for various activities. The specific MET values are as follows: walking = 3.3, moderate-intensity physical activity = 4.0, and high-intensity physical activity = 8.0. The total physical activity index was calculated by summing the METs expended on walking, moderate-intensity physical activity, and high-intensity physical activity (18). According to the WHO physical activity recommendations, individuals whose total MET-minutes per week (MET-min/week) are considered to meet the physical activity standards. Based on the IPAQ classification, participants were divided into two categories: “meeting the standard” (≥600 MET-min/week) and “not meeting the standard” (<600 MET-min/week).

Anxiety levels were assessed using the Generalized Anxiety Disorder 7 (GAD-7) scale. The scale consists of 7 items, each rated on a 4-point Likert scale (0–3 points), with a total score range from 0 to 21 (19). Based on the total score, anxiety levels were categorized as follows: no anxiety (0–4 points), mild anxiety (5–9 points), moderate anxiety (10–14 points), and severe anxiety (15–21 points). The GAD-7 has demonstrated good reliability and validity, and its Chinese version has been validated in Chinese populations (20).

Demographic information of the participants, including age, gender, body mass index (BMI), smoking status (yes/no), alcohol consumption status (yes/no), history of hypertension (yes/no), and history of diabetes (yes/no), was collected via questionnaires. BMI was calculated using the standard formula: BMI = weight (kg) / height² (m²). During the data analysis, these variables were included in the statistical models to control for potential confounding factors, thus reducing the impact of these factors on the relationship between anxiety levels, sleep quality, and physical activity index, thereby enhancing the reliability of the study’s conclusions.

All data were initially recorded by the research team members using paper questionnaires, which were then transcribed into an electronic database. Questionnaire data were exported from the online survey platform, and scores for each indicator were calculated, followed by data organization and cleaning. Data entry and management were performed using Microsoft Excel, and statistical analyses were conducted using SPSS 27.0 (IBM) and GraphPad Prism 8.0.2 (GraphPad Software).

Descriptive statistics were used to summarize the basic characteristics of the study population, with categorical variables presented as frequencies and percentages, and continuous variables expressed as mean ± standard deviation (Mean ± SD) or median (interquartile range, IQR). Before conducting parametric tests, the distribution of continuous variables was assessed for normality using the Shapiro–Wilk test, and homogeneity of variances was examined using Levene’s test. Chi-square tests were used to compare differences between categorical variables, exploring the relationships between different levels of physical activity, sleep quality, and anxiety status. For comparisons of continuous variables between groups, independent t-tests (for normally distributed data) or Mann–Whitney U tests (for non-normally distributed data) were used. Multivariable logistic regression models were employed to assess the associations between physical activity levels, sleep quality, and anxiety status, adjusting for potential confounders. In the regression models, anxiety level was the dependent variable, and physical activity level and sleep quality were the main independent variables, while age, gender, BMI, smoking, and alcohol consumption were included as covariates. The results of the models were presented as odds ratios (OR) with 95% confidence intervals (95% CI), with p < 0.05 considered statistically significant.

A total of 526 volunteers were recruited for this study. After screening, 12 participants with invalid IPAQ questionnaires, 16 participants with incomplete PSQI data, and 10 participants who did not provide anxiety assessments were excluded, resulting in 488 participants included in the final analysis (Figure 1). Based on the GAD-7 scoring criteria, 122 (25.0%) were classified into the anxiety group and 366 (75.0%) into the non-anxiety group.

Table 1 presents the demographic characteristics of the study sample. In the overall sample (n = 488), there were slightly more females than males (52.0% vs. 48.0%), with an average age of 57.82 ± 10.22 years. Most participants resided in urban areas (74.0%), were married (85.0%), and had an education level below high school (83.6%). Regarding lifestyle habits, 65.6% of participants were non-smokers, and 68.0% did not drink alcohol. Regarding health status, 44.5% of participants reported a history of hypertension, 10.9% reported a history of diabetes, and 24.2% were overweight or obese (BMI ≥ 24 kg/m²).

When comparing the demographic characteristics between the anxiety group and the non-anxiety group, no statistically significant differences were found between the two groups in terms of gender (p = 0.917), age (p = 0.217), residential area (p = 0.858), marital status (p = 0.164), education level (p = 0.778), smoking status (p = 0.379), drinking status (p = 0.262), history of hypertension (p = 0.494), history of diabetes (p = 0.208), and BMI (p = 0.272). These results suggest that the demographic distribution of the anxiety and non-anxiety groups in this study was well-balanced, providing a good foundation for subsequent analyses.

Figures 2, 3 show the distributions of sleep quality and physical activity levels between the anxiety and non-anxiety groups. In the anxiety group, the majority of participants (n = 85) exhibited poor sleep quality, while only a few (n = 37) had good sleep quality. In contrast, in the non-anxiety group, most participants (n = 314) had good sleep quality, with a smaller proportion (n = 52) experiencing poor sleep quality. This suggests a close association between anxiety status and sleep quality, with individuals in the anxiety group being more likely to experience sleep problems (Figure 2).

Regarding physical activity levels, only a few participants in the anxiety group (n = 17) met the WHO recommended physical activity standards (≥600 MET-min/week), while most participants (n = 105) did not meet the standard. In the non-anxiety group, although the majority of participants (n = 204) did not meet the recommended physical activity levels, the proportion of those meeting the standard (n = 162) was significantly higher than in the anxiety group. This finding suggests that physical activity levels are also associated with anxiety status, supporting the hypothesis that physical activity may have a protective effect against anxiety (Figure 3).

Univariate logistic regression analysis confirmed significant associations between physical activity levels and anxiety status (p < 0.001) (Table 2). Among all participants (n = 488), 63.3% (309 participants) did not meet the recommended physical activity levels (<600 MET-min/week), while 36.7% (179 participants) met the standard. The proportion of participants in the anxiety group who did not meet the physical activity standard (86.1%) was significantly higher than in the non-anxiety group (55.7%).

For sleep quality, 71.9% (351 participants) had good sleep quality, while 28.1% (137 participants) had poor sleep quality. The proportion of poor sleep quality was markedly higher in the anxiety group (69.7%) than in the non-anxiety group (14.2%) (p < 0.001). Sleep duration did not differ significantly between groups: 6.72 ± 2.21 h in the anxiety group and 6.45 ± 2.16 h in the non-anxiety group (p = 0.172). Poor sleep quality and insufficient physical activity were each strongly associated with higher anxiety risk, whereas sleep duration showed no significant relationship.

To more comprehensively assess the associations between physical activity levels, sleep quality, and anxiety status while controlling for potential confounders, multivariable logistic regression analysis was performed. The analysis included both unadjusted and adjusted models (adjusting for confounders such as age, gender, BMI, smoking, alcohol consumption, hypertension, and diabetes) (Table 3).

In the unadjusted model, compared with poor sleep quality, good sleep quality was associated with a markedly reduced risk of anxiety (OR = 0.069, 95% CI: 0.041–0.115, p < 0.001). After adjusting for potential confounders, the association remained significant and similar in magnitude (OR = 0.065, 95% CI: 0.038–0.111, p < 0.001).

For physical activity, in the unadjusted model, not meeting the WHO recommended standard was associated with substantially higher anxiety risk compared with meeting the standard (OR = 5.345, 95% CI: 2.849–10.027, p < 0.001). This relationship persisted after adjustment (OR = 5.408, 95% CI: 2.827–10.344, p < 0.001).

Both good sleep quality and meeting WHO physical activity standards were independently associated with lower anxiety risk, and these relationships remained robust after controlling demographic and lifestyle factors.

To explore whether the effects of physical activity levels and sleep quality on anxiety differ across various demographic characteristics and health conditions, stratified analyses were conducted.

In the stratified analysis of sleep quality, it was found that the effect of sleep quality on anxiety was significant in all subgroups of demographic characteristics and health conditions (p < 0.05), with no statistically significant interactions (all P-interaction > 0.05). In the female subgroup, the protective effect of good sleep quality on anxiety (OR = 21.248, 95% CI: 10.382–43.487, p < 0.001) appeared numerically stronger than in the male subgroup (OR = 9.219, 95% CI: 4.725–17.986, p < 0.001). However, the interaction was not statistically significant (P-interaction = 0.095), and therefore, this observation should be interpreted with caution as it was not confirmed statistically. For residential area, the ORs were 14.472 (95% CI: 8.220–25.479, p < 0.001) in urban residents and 12.300 (95% CI: 4.793–31.567, p < 0.001) in rural residents; While the point estimate was slightly higher in urban participants, no statistically significant interaction was detected (P-interaction = 0.772), and the observed difference should not be overinterpreted. In the marital status subgroup, married individuals (OR = 16.099, 95% CI: 9.355–27.705, p < 0.001) showed a higher point estimate compared with those of other marital statuses (OR = 7.086, 95% CI: 2.346–21.402, p < 0.001), but the difference was not statistically significant (P-interaction = 0.192). No significant differences were found in the education level subgroup (P-interaction = 0.690), with non-smokers (OR = 17.539, 95% CI: 9.525–32.293, p < 0.001) and non-drinkers (OR = 18.070, 95% CI: 9.911–32.945, p < 0.001) showing stronger protective effects of sleep quality on anxiety compared to smokers (OR = 8.898, 95% CI: 3.967–19.959, p < 0.001) and drinkers (OR = 7.744, 95% CI: 3.343–17.942, p < 0.001), although the interaction was not statistically significant (Figure 4).

In the stratified analysis of physical activity levels, the effect of physical activity on anxiety was significant in all subgroups (p < 0.05), with no statistically significant interactions (all P-interaction > 0.05). Unlike sleep quality, the effect of physical activity on anxiety did not differ significantly between males (OR = 0.191, 95% CI: 0.085–0.425) and females (OR = 0.217, 95% CI: 0.101–0.464) (P-interaction = 0.819). In the residential area subgroup, the protective effect of physical activity on anxiety was stronger in urban residents (OR = 0.157, 95% CI: 0.078–0.315, p < 0.001) compared to rural residents (OR = 0.359, 95% CI: 0.141–0.914, p = 0.032), but the interaction was not statistically significant (P-interaction = 0.163). In terms of marital status, the protective effect of physical activity on anxiety was stronger in married individuals (OR = 0.222, 95% CI: 0.124–0.395, p < 0.001) compared to those in other marital statuses (OR = 0.106, 95% CI: 0.013–0.86, p = 0.036), with no significant interaction (P-interaction = 0.506). No significant differences were observed in the education level, smoking status, drinking status, hypertension history, diabetes history, or BMI subgroups (all P-interaction > 0.05). Among non-drinkers, meeting physical activity standards had a stronger protective effect on anxiety (OR = 0.25, 95% CI: 0.131–0.474, p < 0.001) compared to drinkers (OR = 0.129, 95% CI: 0.043–0.388, p < 0.001). The protective effect of physical activity on anxiety was stronger in those with normal weight (OR = 0.163, 95% CI: 0.081–0.329, p < 0.001) compared to overweight or obese individuals (OR = 0.314, 95% CI: 0.123–0.8, p = 0.015), although these differences did not reach statistical significance (Figure 5).

The protective effects of good sleep quality and sufficient physical activity were consistent across demographic and health subgroups, and no statistically significant interactions were observed.