Fatigue, anthropometric profile, and health-related quality of life among fifth-year medical students: A single-center cross-sectional study from Kazakhstan

Abstract

BACKGROUND:

Fatigue is a familiar problem in medical training, and a growing body of work has linked it to worse academic performance and lower psycho-emotional well-being. Few data are available for Central Asian student cohorts. We set out to measure the prevalence of fatigue in a single fifth-year cohort and to test how it relates to body mass index (BMI), physical activity, and self-reported quality of life (QoL).

MATERIALS AND METHODS:

Cross-sectional, single-center survey conducted in the middle of the academic semester at a medical faculty in Kazakhstan, on fifth-year students only. Two hundred and sixty-nine fifth-year medical students completed the survey. We measured fatigue with the fatigue assessment scale (FAS) and the fatigue severity scale (FSS), and quality of life with the Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors (COOP-WONCA) charts. Body mass index and physical activity (short international physical activity questionnaire [IPAQ] adaptation) were collected by self-report. Descriptive statistics, one-way analysis of variance (ANOVA), the Chi-square test, and multinomial logistic regression were run in IBM SPSS Statistics 23.0. Two-tailed P < 0.05.

RESULTS:

Clinically significant fatigue (FSS ≥ 4) was found in 27.7% of the cohort. The composite COOP-WONCA score split students into three groups (good, moderate, poor quality of life). Physical inactivity was the strongest independent predictor of falling into the poor-QoL group (OR ≈ 3.4–3.6, P < 0.05), while higher FSS scores independently predicted moderate (vs good) QoL (OR = 1.05 per point, P = 0.048). BMI was not an independent predictor.

CONCLUSIONS:

Roughly 1 in 4 fifth-year medical students reach the clinical threshold for fatigue by mid-semester, and this is associated with measurable losses in daily and social functioning. Activity-level interventions look like the most promising entry point. To our knowledge, this is the first Central Asian study to combine FAS, FSS, and COOP-WONCA in one cohort and use multinomial logistic regression to identify independent predictors of poor QoL in medical students.

Introduction

Fatigue is a normal feature of any living organism.[1] In primary care, though, it is anything but trivial: over 10% of patients seen by general practitioners list significant fatigue among their complaints.[2] The symptom is non-specific. It shifts with age, sex, physical condition, diet, and psychological state that is exactly what makes the diagnosis tricky.[3] Fatigue is one of the most frequent subjective complaints in clinical practice, and it has physical, cognitive, and emotional dimensions that, together, predict later health problems and interfere with daily life.[4] The picture is sharper in medical schools, where academic stress feeds into burnout and pulls quality of life (QoL) down.[5] A tired body also tends to be more vulnerable, with cardiovascular and immune disturbances among the most frequently described consequences.[6] Why does this matter for the medical profession in particular? Because fatigue distorts the very things future clinicians are being trained for: the accuracy of manual procedures and the perception of applied force.[7] Tools for measuring fatigue range from simple visual-analogue scales[8] to multidimensional questionnaires.[9] For non-clinical student samples, however, only validated instruments are recommended. Two of them, used here, are the fatigue severity scale (FSS)[10] and the fatigue assessment scale (FAS).[11] Our aim was to see how FAS and FSS scores relate to anthropometric data and QoL in fifth-year medical students, so that risk factors for poor well-being might be picked up earlier.

Novelty of the study. As far as we are aware, no previous study from Central Asia has put FAS, FSS, and the Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors (COOP-WONCA) charts together in a fifth-year medical-student cohort and then used multinomial logistic regression to pull out independent predictors of poor QoL. Most existing work from neighboring regions has settled for one fatigue scale, or skipped multivariable analysis, or has been timed to exam periods, when fatigue is at its peak. By running our survey in the middle of the semester, we step around the acute exam-stress confound and capture something closer to a baseline.

Theoretical framework. Our working model is a simple mediator one [Figure 1]. Two measured predictor groups, anthropometric profile (body mass index [BMI]) and lifestyle (physical activity), are assumed to push into fatigue. Fatigue here is captured by FAS (mental and general aspects) and FSS (the functional impact on daily life). Fatigue, in turn, is read as the proximal driver of reduced health-related QoL across the COOP-WONCA domains. Because our design is cross-sectional, we do not infer the direction of causality. The dashed arrow in Figure 1 marks the reverse path, the likely reciprocal effect of low QoL back on the predictor side.

Figure 1.

Theoretical framework of the study. Two groups of measured predictors—anthropometric profile (BMI) and lifestyle factors (physical activity)—feed into fatigue (captured by FAS and FSS) that in turn mediates the impact on health-related quality of life across the COOP-WONCA domains. The dashed arrow indicates the reciprocal influence of reduced quality of life on the predictor side; direction of causality cannot be inferred from the present cross-sectional design. BMI = Body mass index; COOP-WONCA = Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors; FAS = Fatigue assessment scale; FSS = Fatigue severity scale

Materials and Methods

Study design and setting

This was a single-center, cross-sectional observational study run at a medical faculty in Kazakhstan. We deliberately chose mid-semester for data collection. The reason was simple: end-of-semester or pre-exam surveys catch acute stress that inflates fatigue scores and distorts any cross-sectional reading of baseline well-being. Mid-semester gets us closer to a steady state.

Study participants and sampling

Every fifth-year student attending the faculty during the survey period was invited to take part. The inclusion criteria were: written informed consent, current enrolment in the final year, and no current exacerbation of a chronic somatic condition. Students who declined the full questionnaire battery or did not meet these criteria were not entered. Out of the initial pool, 56 were excluded—either for incomplete responses or for an active illness on the day of the survey. That left 269 in the analytic sample. Sampling was non-probability, by convenience that is consistent with a single-center design.

Data collection tool and technique

All instruments were administered in Kazakh. The FSS[10] is a nine-item questionnaire on a seven-point Likert scale; it captures how much fatigue interferes with day-to-day and social activity. Published internal consistency for the FSS is good (α = 0.79–0.96) and has been re-confirmed in subsequent validations.[12] The FAS[11] has ten items on a five-point Likert scale. Unlike older fatigue measures, it was built to separate the mental from the physical and to keep somatic-symptom contamination low that is why it works well in non-clinical student samples.[13] For quality of life we used the COOP-WONCA charts,[14] covering six domains: physical condition, feelings, daily activities, social activities, change in health status, and overall health. Physical activity was measured with a short adaptation of the international physical activity questionnaire (IPAQ), and anthropometric data (age, sex, height, weight, BMI) came from a standard self-report form.

Ethical consideration

The study was approved by the institutional research ethics committee at the host university (Protocol No. 9 dated 15 November 2022). Every participant received an information sheet and signed a written consent form before completing any questionnaires. Participation was voluntary, and students could withdraw at any point with no consequence. All data were anonymized at entry. Where FAS or FSS scores indicated clinically significant fatigue, the student was proactively offered a referral to the faculty’s student-health and counselling service.

Statistical analysis

Analyses were performed in IBM SPSS Statistics, version 23.0. Continuous variables are reported as mean ± SD; categorical variables as counts and percentages. For the continuous variables (age, FAS sum, FSS sum), we used one-way ANOVA across the three COOP-WONCA QoL groups (good, moderate, poor). For the categorical variables (sex, BMI category, physical activity level, FSS ≥ 4), we used the Chi-square (χ²) test, both across the QoL groups and, in a separate set, to compare COOP-WONCA domain responses between fatigue strata (FSS < 4 vs FSS ≥ 4).

For the multivariable step, we fitted multinomial logistic regression with the composite COOP-WONCA category (good/moderate/poor, derived from the sum score) as the dependent variable. Two models were run in parallel: Model 1 with FSS sum as the principal exposure, Model 2 with FAS sum. Both were adjusted for BMI, sex (reference: female), and physical activity (reference: regular activity). The good-QoL category served as the reference outcome. Effects are reported as odds ratios (OR) with 95% confidence intervals. All tests were two-tailed; P < 0.05 was the threshold for significance.

Results

After screening, 56 participants were excluded, and 269 fifth-year medical students remained in the analytic sample. The mean age was 21.9 ± 2.7 years. We used the Kazakh-language versions of the FAS, FSS, IPAQ, and COOP-WONCA charts.

Baseline characteristics and fatigue burden across COOP-WONCA groups

Fatigue was common in this cohort. Twenty-seven-point seven percent (n = 74) of students were at or above FSS ≥ 4, the clinically significant cut-off. Based on the composite COOP-WONCA score, the cohort split into three QoL groups: good QoL (n = 148; 55.0%), moderate QoL (n = 86; 32.0%), and poor QoL (n = 35; 13.0%). Baseline characteristics across these groups are shown in Table 1. Mean FAS and FSS scores stratified by sex, physical-activity level, and BMI category are shown in [Figure 2].

Table 1.

Baseline characteristics and fatigue burden of fifth-year medical students stratified by COOP-WONCA quality-of-life group (n=269)

Variable		Total (n=269)		Good QoL (n=148)		Moderate QoL (n=86)		Poor QoL (n=35)		P
Age, years, M (SD)		21.9 (2.7)		21.7 (2.5)		22.1 (2.8)		22.3 (3.0)		0.310
Sex, n (%)										<0.001
Male		108 (40.1)		78 (52.7)		16 (18.6)		14 (40.0)
Female		161 (59.9)		70 (47.3)		70 (81.4)		21 (60.0)
BMI category, n (%)										0.182
Underweight (<18.5)		27 (10.0)		14 (9.5)		9 (10.5)		4 (11.4)
Normal (18.5–24.9)		188 (69.9)		108 (73.0)		56 (65.1)		24 (68.6)
Overweight/Obesity		54 (20.1)		26 (17.6)		21 (24.4)		7 (20.0)
Physical activity, n (%)										<0.001
None		67 (24.9)		25 (16.9)		25 (29.1)		17 (48.6)
<150 min/week		121 (45.0)		65 (43.9)		40 (46.5)		16 (45.7)
≥150 min/week		81 (30.1)		58 (39.2)		21 (24.4)		2 (5.7)
Fatigue indicators
FAS sum, M (SD)		14.0 (4.7)		11.8 (3.5)		14.5 (4.2)		17.6 (5.1)		<0.001
FSS sum, M (SD)		3.1 (1.3)		2.4 (0.7)		3.5 (1.1)		5.1 (1.4)		<0.001
FSS ≥ 4, n (%)		74 (27.7)		12 (8.1)		30 (34.9)		32 (91.4)		<0.001

M=mean; SD=standard deviation; P=significance level. For continuous variables (age, FAS, FSS), P is from one-way ANOVA across the three QoL groups. For categorical variables (sex, BMI category, physical activity, FSS ≥ 4), P is from the chi-square (χ²) test. Bold P values denote statistical significance (P < 0.05). BMI=Body mass index; COOP-WONCA=Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors; FAS=Fatigue assessment scale; FSS=Fatigue severity scale

Figure 2.

Mean FAS and FSS scores by (a) sex, (b) physical activity level, and (c) body mass index category among fifth-year medical students (n = 269). Error bars represent ± 1 standard deviation. Brackets with asterisks denote statistically significant ANOVA contrasts (P < 0.05). The dotted overlay on panel (c) highlights the J-shaped association between BMI category and fatigue scores. FAS = Fatigue assessment scale; FSS = Fatigue severity scale; BMI = Body mass index; ANOVA = Analysis of variance

A few patterns stand out. The proportion of women rose sharply with worsening quality of life: 47.3% in the good-QoL group, 81.4% in the moderate-QoL group (P < 0.001 across groups). Physical inactivity tracked the same gradient. Regular activity (≥150 min/week) was reported by 39.2% of students with good QoL, but by only 5.7% of those with poor QoL (P < 0.001). Both fatigue scores rose monotonically across groups: mean FAS climbed from 11.8 to 17.6, and mean FSS from 2.4 to 5.1 (both P < 0.001). The proportion meeting the clinical FSS ≥ 4 threshold went from 8.1% in the good-QoL group to 91.4% in the poor-QoL group. BMI categories did not differ significantly across QoL groups (P = 0.182), and mean age was essentially the same in all 3 (21.7–22.3 years; P = 0.310).

Quality-of-life domains across fatigue groups

QoL responses on individual COOP-WONCA domains also differed between fatigue strata. Chi-square testing flagged significant differences in several domains (P < 0.05).

Daily activity. Fatigued students were less physically active. Among those with FSS < 4, 45.5% reported high physical activity; in the FSS ≥ 4 group only 31.9% did (χ² = 10.03, P = 0.04).

Health assessment. Self-rated health was lower in the fatigued group. Only 20% of fatigued students rated their health as excellent, compared with more than a third of non-fatigued students (χ² = 10.41; P = 0.032).

Physical fitness. High self-reported fitness was less frequent in the FSS ≥ 4 cohort (15.2%) than in the FSS < 4 cohort (30.8%) (χ² = 9.06; P = 0.03).

Multivariable analysis (multinomial logistic regression)

To get at independent predictors of poor quality of life, we fitted two multinomial logistic regression models. The dependent variable was the COOP-WONCA composite (good/moderate/poor) and the good-QoL category served as the reference outcome. Model 1 used FSS sum as the principal exposure, Model 2 used FAS sum. Both models adjusted for BMI, sex (reference: female), and physical activity level (reference: regular activity). Full results are in Table 2.

Table 2.

Multinomial logistic regression: Predictors of COOP-WONCA quality-of-life category (reference outcome: Good QoL)

Predictor		Moderate vs. Good QoL				Poor vs. Good QoL
		OR (95% CI)		P		OR (95% CI)		P
Model 1—Fatigue Severity Scale (FSS)
FSS sum (per 1-point increase)		1.05 (1.00–1.10)		0.048		1.03 (0.98–1.07)		0.270
BMI (per 1 kg/m²)		1.09 (0.94–1.26)		0.254		1.11 (0.96–1.28)		0.165
Male sex (vs. female)		0.18 (0.06–0.52)		0.002		1.05 (0.38–2.88)		0.925
Low physical activity (vs. regular)		2.56 (0.95–6.90)		0.064		3.56 (1.30–9.78)		0.014
Model 2—Fatigue Assessment Scale (FAS)
FAS sum (per 1-point increase)		1.08 (0.96–1.20)		0.192		0.99 (0.88–1.11)		0.821
BMI (per 1 kg/m²)		1.10 (0.95–1.27)		0.209		1.12 (0.97–1.30)		0.117
Male sex (vs. female)		0.21 (0.07–0.59)		0.003		1.15 (0.42–3.11)		0.791
Low physical activity (vs. regular)		2.23 (0.84–5.92)		0.109		3.38 (1.24–9.17)		0.017

OR=odds ratio, CI=confidence interval. Reference for sex: female. Reference for physical activity: regular activity. Statistically significant P values (<0.05) are highlighted in bold. BMI=Body mass index; COOP-WONCA=Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors; FAS=Fatigue assessment scale; FSS=Fatigue severity scale; QoL=Quality of life

After adjustment, the FSS sum was an independent predictor of moderate (vs good) quality of life: OR = 1.05 per point (95% CI 1.00–1.10; P = 0.048). The FAS sum did not reach significance as an independent predictor in either contrast.

Physical inactivity came out as the strongest independent predictor of poor (vs good) quality of life. In Model 1, OR = 3.56 (95% CI 1.30–9.78; P = 0.014); in Model 2, OR = 3.38 (95% CI 1.24–9.17; P = 0.017). Male sex was protective against falling into the moderate-QoL category (vs good QoL): OR = 0.18 in Model 1 (P = 0.002) and OR = 0.21 in Model 2 (P = 0.003). It made no difference, however, for poor QoL. BMI did not reach significance in any contrast, which suggests that the bivariate BMI-fatigue link visible in Table 1 is mostly carried by lifestyle factors rather than by body composition itself.

Figure 3 makes the picture visible briefly. For the poor-QoL contrast, the entire confidence interval for physical inactivity sits to the right of the null line in both models that is what one would expect from the strongest predictor. For the moderate-QoL contrast, male sex pulls the CI well to the left of the null, consistent with a protective association. The BMI estimates cross the null line in every contrast that is what led us to read the bivariate BMI-fatigue link in Table 1 as largely mediated by physical activity.

Figure 3.

Forest plot of multinomial logistic regression results (n = 269). Two parallel models are presented: Model 1 with the FSS sum score as principal exposure and Model 2 with the FAS sum score. Each model is adjusted for BMI, sex (reference: female) and physical activity level (reference: regular activity). The reference outcome is good quality of life. Saturated colors mark statistically significant odds ratios; faded colors mark non-significant estimates. Horizontal bars represent 95% confidence intervals; markers denote point estimates. The dashed vertical line indicates the null effect (OR = 1). FAS = Fatigue assessment scale; FSS = Fatigue severity scale; BMI = Body mass index; OR = Odds ratio

Discussion

Clinically significant fatigue (FSS ≥ 4) was present in 27.7% of our fifth-year students (n = 269; mean age 21.9 ± 2.7 years). For reference, Yoon’s meta-analysis gives a general-population prevalence of 20–23%,[15] and our figure sits above that, which is what one would expect in a medical-training context. Bouloukaki showed a sharp climb in FSS scores during exam periods,[16] and two more recent papers report rates far higher than ours: Alghamdi found 88.6% on FAS[17] and Hanif found 61.2%.[18] The gap between our 27.7% and Alghamdi’s 88.6% almost certainly reflects timing. We ran the survey mid-semester, on purpose; those two studies caught their participants at the academic peak. The contrast also hints that FAS is more sensitive than FSS to milder forms of exhaustion.

Physical inactivity tracked closely with higher FAS scores in our cohort: mean FAS was 16.95 in the inactive group versus 11.20 in the active group (P = 0.038). Alghamdi reported the same association at the same significance level (P = 0.038),[17] as did Hanif (P < 0.05).[18]

Li and Lalović reach a similar conclusion about exercise as a buffer against academic fatigue.[19,20] The pattern is also closely echoed by a JEHP study from Tabriz, where Ghassab-Abdollahi found that physical activity and depression together explained about 63% of the variance in QoL scores among 186 medical students assessed with IPAQ and WHOQOL[21]—strikingly close to what we see here. Marijančić et al., working with a Croatian student cohort, named sedentary behavior as the main fatigue risk factor regardless of sex,[22] and Wittayapun et al. picked up the same trend in the post-COVID setting.[23]

Body mass index and fatigue

Overweight students had significantly higher fatigue scores in our bivariate analysis: FAS = 16.76 (P = 0.041) and FSS = 6.02 (P = 0.038). Underweight students also crossed the FSS threshold (4.62), so the pattern across BMI categories is more J-shaped than linear. Hoseini treats BMI as a mediator between physical activity and QoL[24] that fits well with our reading and with Lalović.[20] The picture is also in line with Szemik[25] and Dejhalla, who reported metabolic syndrome in 5.5% of students and singled out obesity and low physical activity as the main risk factors.[26]

QoL worsened in step with fatigue. Mean FAS climbed from 11.8 in the good-QoL group to 17.6 in the poor-QoL group, and the proportion meeting the clinical FSS ≥ 4 threshold went from 8.1 to 91.4%—a steep, ordered gradient (P < 0.001 in both cases). The meta-analysis by Abrantes, drawing on almost 20,000 students, points in the same direction.[27] Liu and colleagues likewise found academic burnout to be both common and associated with reduced well-being in university students.[28] In our data, FAS appears to track psycho-emotional load, while FSS captures the more functional, physical side.[29] Both moved together with the overall COOP-WONCA composite.

Both instruments come with a long psychometric record. Muranaka et al. showed strong measurement properties for the FSS,[30] and Naik et al. confirmed its construct validity in patients recovering from COVID-19, with strong correlations to QoL measures.[12] For the FAS, the review by Hendriks et al. summarized its psychometric quality and the availability of translated versions across multiple disease populations.[13]

Strengths and limitations

Strengths. To our knowledge, this is the first Central Asian study to bring three validated instruments (FAS, FSS, COOP-WONCA) together in a single fifth-year medical-student cohort and then use multinomial logistic regression to separate independent predictors of poor QoL. A second strength is the timing: by collecting mid-semester, we step around the well-known inflation of fatigue scores by acute exam stress. Third, the sample (n = 269) is adequate for a four-predictor logistic model and comfortably above the events-per-variable threshold needed for stable estimates.

Limitations. Several caveats apply. (i) The design is single-center and includes only fifth-year students, which limits generalizability to other training years and to other institutional contexts. (ii) Physical activity and anthropometric data were self-reported, which introduces social-desirability and recall bias. (iii) The cross-sectional design rules out causal claims about fatigue and QoL. (iv) Several plausible confounders were not measured and so could not be adjusted for: sleep quality, anxiety, depression, caffeine intake, weekly study hours, paid part-time work, and socio-economic status. The adjusted estimates for FSS and physical activity should therefore be read as conditional only on the four variables modelled. (v) We did not re-estimate Cronbach’s α in our own sample and relied on published values. (vi) A formal response rate could not be computed, because the total number of eligible students invited was not recorded. (vii) Although Chi-square testing was used for individual COOP-WONCA domains, we did not apply a formal correction for multiple comparisons; P values for those domain-level analyses should be interpreted with that caveat in mind.

Directions for further research

Several extensions are worth pursuing. A longitudinal cohort that follows students across pre-clinical, clinical, and postgraduate years would map fatigue trajectories against the major academic and life transitions and support causal inference. A multi-center design across several medical faculties—ideally in different countries of Central Asia and beyond—would address the generalizability concerns. Future surveys should bring in the missing variables (sleep, anxiety, depression, caffeine, study hours, socio-economic indicators) so that adjusted models can isolate the contribution of each lifestyle factor more cleanly.

Interventions are the natural next step. Structured physical-activity programmers, sleep-hygiene workshops, and curriculum-level stress-reduction modules should be tested in cluster-randomized designs, with FAS, FSS, and COOP-WONCA scores as primary endpoints.

Conclusion

In a cohort of 269 fifth-year medical students (mean age 21.9 ± 2.7 years), the headline finding is that 27.7% already had clinically significant fatigue by mid-semester. Bivariate analyses tied higher BMI and lower physical activity to higher fatigue scores. Multinomial logistic regression then confirmed two independent predictors of poor quality of life: physical inactivity (OR ≈ 3.4–3.6) and, in the FSS model, fatigue severity itself (OR = 1.05 per point). Female sex was associated with higher odds of falling into the moderate-QoL band, rather than remaining in the good-QoL band.

On the COOP-WONCA charts, fatigue showed up across every domain of daily life, with the heaviest effects in “daily activities” and “social activities.” Students with high FAS and FSS scores rated these areas as noticeably reduced, and they also rated their physical and psycho-emotional state lower overall. Fatigue in this population is, in other words, tied to specific lifestyle parameters and produces a measurable drop in QoL.

Implications for health policy and medical education

Our findings have practical implications for institutional policy and for how the medical curriculum is built:

1. Routine screening. Medical faculties could fold FAS and FSS (together with the COOP-WONCA charts) into the annual student-health review, alongside the usual physical and mental-health checks. Mid-semester rather than end-of-semester is the safer window, for reasons set out above.

2. Physical activity in the curriculum. Since physical inactivity was the single strongest independent predictor of poor quality of life in our data, faculties should build structured opportunities for regular activity into the academic schedule—protected time for fitness, on-campus sport facilities open during clinical placements, and academic credit for documented weekly activity.

3. A targeted counselling pathway. Students who score FSS ≥ 4 on routine screening should be offered a short, low-threshold consultation with the student-health service. That consultation can also screen for sleep disturbance, anxiety, and depression—variables our data could not capture but that are well documented in similar cohorts—and refer onwards as needed.

4. Curriculum content on fatigue and self-care. Because fatigue distorts manual skill performance and clinical decision-making, the fifth-year curriculum should include an explicit module on recognizing, managing, and disclosing fatigue. It can cover sleep hygiene, physical-activity guidelines, time-management, and the professional duty to flag fatigue during clinical work.

5. Institutional monitoring. Faculties should track aggregate FAS, FSS, and COOP-WONCA indicators each year as a quality marker of the educational environment, in line with established practice for monitoring burnout in postgraduate medical training.

Together these steps move beyond passive recognition of student fatigue. They give medical schools a feasible, structured framework that fits within the student-wellness services already in place.

Use of artificial-intelligence tools

In line with the policy of the Journal of Education and Health Promotion and the ICMJE recommendations, we disclose the following use of artificial-intelligence tools during the preparation of this manuscript. An AI-based writing assistant (Claude, Anthropic) was used for English-language editing, for formatting references in Vancouver style, and for rendering figures from our own data. The AI tool was not used to design the study, to collect, analyses, or interpret the underlying data, or to generate scientific claims or conclusions. The study design, data acquisition, statistical analysis, and clinical interpretation were carried out by the authors. The authors take full responsibility for the content, scientific accuracy, and integrity of the work, including every section in which AI-assisted editing was applied.

Abbreviations

ANOVA: analysis of variance; BMI: body mass index; CI: confidence interval; COOP-WONCA: Dartmouth COOP Functional Health Assessment Charts/World Organization of Family Doctors; FAS: Fatigue Assessment Scale; FSS: Fatigue Severity Scale; ICMJE: International Committee of Medical Journal Editors; IPAQ: International Physical Activity Questionnaire; IRB: Institutional Review Board; MeSH: Medical Subject Headings; OR: odds ratio; QoL: quality of life; SD: standard deviation.

Ethical approval

The study was approved by the Institutional Research Ethics Committee of Khoja Akhmet Yassawi International Kazakh-Turkish University, Turkestan, Kazakhstan (Protocol No. 9, dated 15 November 2022). The study was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to their inclusion; participation was voluntary and could be withdrawn at any time without consequence.

Author’s contributions

DI and KS conceived and designed the study. DI, SK and AA collected the data. DI and SK performed the statistical analysis and interpreted the results. DI drafted the manuscript. KS and AA critically revised it for important intellectual content. KS supervised the project and is the corresponding author. All authors read and approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Data availability statement

The anonymized datasets generated and analyzed during the current study are not publicly available because participant consent did not cover open data deposition, but they are available from the corresponding author (K. Sadykova, karlygash.sadykova@ayu.edu.kz) on reasonable request and subject to approval by the institutional research ethics committee.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.