Aerobic exercise to alleviate primary dysmenorrhea in adolescents and young women: A systematic review and meta‐analysis of randomized controlled trials

Abstract

Introduction

Aerobic exercise has been confirmed to alleviate primary dysmenorrhea (PD) in adolescents and young women. However, the effect of the aerobic exercise type and dosage on PD alleviation was unclear. This research aims to assess the effect of aerobic exercise on PD and investigate the dose–response relationships.

Material and Methods

Systematic literature searches of Web of Science, Embass, Cochrane Library, PubMed, PsycNET, CINAHL, CNKI, Baidu Scholar, Google Scholar, and other Complimentary Medicine Database. PICOS standards were adopted in this research: participants were nonathlete women with PD; intervention was aerobic exercise of at least one menstruation cycle; comparator was any comparator; outcomes were pain intensity or pain duration; and study type was randomized controlled trials. The Cochrane Collaboration risk of bias tool was used to assess the quality of the research. Random‐effect meta‐analysis was conducted for pain intensity and pain duration, with prespecified subgroup analyses based on aerobic exercise components. The strength of the evidence was assessed using GRADE. This systematic review and meta‐analysis was registered in PROSPERO (CRD42024533544).

Results

The study identified 16 eligible studies, with 15 involving adolescents, totaling 918 participants aged 15 to 43, with an average age of 21.26 ± 13.15. The results confirmed that aerobic exercise can alleviate PD's pain intensity (standard mean difference (SMD) = −1.728 (p = 0.00), 95% CI [−2.26 to −1.31]) and pain duration (weighted mean difference (WMD) = −12.53 h, p = 0.01, 95% CI: −21.38 to −3.68). However, the heterogeneity of these two results was high. Subgroup analysis showed that Pilates (SMD = −3.17, 95% CI [−4.26 to −2.07]), low intensity (SMD = −1.64, 95% CI [−2.10 to −1.19]), 31–45 min duration (SMD = −3.05, 95% CI [−5.36 to −0.75]), ≤2 times per week frequency (SMD = −2.24, 95% CI [−3.36 to −1.12]), and a period cycle of 2 menstrual cycles (SMD = −2.21, 95% CI [−3.13 to −1.28]) had the maximum effect size.

Conclusions

Aerobic exercise was able to alleviate pain intensity and pain duration in adolescents and young women with PD. Moderate‐quality evidence indicates that Pilates, low intensity, 46–60 min, ≤two times per week, or two menstrual cycles showed more efficiency in alleviating PD. Due to the limited data, future research should prioritize conducting randomized controlled trials of aerobic exercise interventions in younger age groups to develop personalized treatment strategies for adolescents.

The results of this study indicate that different types and dosages of aerobic exercise produce varying dose‐response effects on pain intensity and duration in primary dysmenorrhea. It is essential to develop more precise aerobic exercise prescriptions for adolescents and young women within the specific clinical context.

Abbreviations

PD

primary dysmenorrhea

RCT

randomized controlled trial

SMD

standard mean difference

WMD

weighted mean difference

Key points.

Aerobic exercise was able to alleviate pain intensity and pain duration in adolescents and young women with primary dysmenorrhea. Some aerobic exercise dose components (such as type, intensity, duration, frequency, and period) influenced the alleviation of aerobic exercise on primary dysmenorrhea.

1. INTRODUCTION

Primary dysmenorrhea (PD) is defined as cramping pain occurring before or during a period in the absence of pelvic pathology. ¹ , ² It is usually diagnosed in adolescents and young women with the symptoms of headache, nausea, and fatigue. ³ , ⁴ The prevalence of PD was 71% in young women and up to 93% in adolescents, ⁴ , ⁵ with over 50% of them reporting moderate or severe symptoms. ⁶ Many researches have confirmed that PD could lead to mood disturbance, ⁷ daily activities decrease, ⁸ absence from school or company, ⁵ , ⁹ and any other problems, with a sizeable economic burden to individuals, education, and society. ⁵ , ⁶ , ⁹ , ¹⁰ Tanaka et al. reported that the financial burden of medical outpatient service, consumption of over‐the‐counter, and reduction in productivity resulting from dysmenorrhea of women exceeds $8.6 billion. ¹¹ The primary treatment of this disease was mainly by nonsteroidal anti‐inflammatory drugs like mefenamic acid, ⁴ , ¹⁰ but the drugs may cause side effects such as stomach irritation, menstrual disorder, or venous thrombus. ¹ , ⁹ , ¹² Moreover, 25% of women refused to take medicines due to religion, health, pregnancy preparation, or cost. ¹³ Thus, more safe, economical, and non‐medical measures to alleviate PD are urgently needed.

Previous research has shown that aerobic exercises such as brisk walking, Pilates, dance, and yoga can effectively alleviate PD. ¹³ , ¹⁴ , ¹⁵ It also has many advantages including low skills requirements, ease of regulation, and many health benefits. ¹⁶ , ¹⁷ , ¹⁸ Physiologically, aerobic exercise can enhance blood circulation in the pelvic cavity, increase the pain threshold, and reduce inflammatory response, alleviating pain. ¹⁰ , ¹⁸ , ¹⁹ This was achieved by activating prostaglandin synthetic inhibitors, stimulating the brain to release analgesic drugs, and acting on mediators of inflammation. ¹⁹ Notably, aerobic exercise's different types or dose components (such as frequency, intensity, duration, and period) do not have a proportional effect on alleviating pain. ²⁰ Furthermore, some studies have also found that excessive exercise may worsen menstrual pain. ¹⁶ , ²¹ Hence, exploring the relationship between aerobic exercise's type or dose components in alleviating PD is necessary.

Over the last 5 years, five meta‐analyses on related topics have been published in peer‐reviewed journals. These studies stated the efficacy and safety of exercise intervention for PD and provided important evidence for clinical decision–making. ² , ⁹ , ²² , ²³ , ²⁴ However, due to the limitations of the included studies in eligibility criteria, low‐quality evidence, and high heterogeneity levels, these results varied and needed to be interpreted cautiously. ⁹ No research investigates how different exercise dosage levels influence pain alleviation. This limitation hampers the ability to draw conclusions regarding the optimal exercise intervention dosage. Therefore, considering the manifold advantages of aerobic exercise and its readily available dosage parameters, this study employs meta‐analysis to investigate potential moderating effects associated with aerobic exercise dose components (including type, intensity, frequency, duration, and menstrual cycle). In addition, subgroup analyses were conducted to account for heterogeneity and assess general trends regarding dose–response relationships, aiming to provide specific guidance and recommendations regarding the optimal dosage for aerobic exercise interventions targeting women with PD.

2. MATERIAL AND METHODS

2.1. Protocol and registration

This systematic review was conducted according to the Cochrane Handbook and was reported following the PRISMA statement. ²⁵ , ²⁶ The prospective registration number is CRD42024533544.

2.2. Search strategy

Two researchers conducted systematic literature searches of Web of Science, Embass, Cochrane Library, PubMed, PsycNET, CINAHL, CNKI, and other Complimentary Medicine Database. The search ranged from establishing the databases to 2024; the final search date was March 1, 2024. The MeSH terms and free words were used for search, including “Dysmenorrhea (MeSH),” “Exercises (MeSH),” “Menstrual Pain,” and “Physical Activity.” No language, date, or publication type restrictions were applied to ensure the comprehensiveness of the references, and no randomized controlled trial (RCT) filters were used. ²⁷ Detailed search information is provided in the Supplementary Information (Appendix S1).

2.3. Eligibility criteria

Published studies in any language were included where the following PICOS (participants [P], intervention [I], comparisons [C], outcomes [O], and study type [S]) criteria were met:

• P: nonathlete females with regular menstruation, experiencing moderate or severe PD, not using hormonal contraception.

• I: aerobic exercise interventions delivered over one menstrual cycle, as a single intervention or as a cointervention, in any setting and via any mode of delivery.

• C: any comparator that did not involve aerobic or other exercise, including active comparators and usual care or no treatment.

• O: Pain intensity is measured by a validated tool, or pain duration is measured in hours.

• S: RCTs.

The following circumstances would be excluded: quasi‐experimental studies, systematic review, or the second analysis; subjects involving regularly exercised individuals, athletes, disabled, or pregnant women; studies with a high rate of participant dropout or lacking pain intensity as an outcome measure or involving unqualified comparators; incomplete or unextractable data; significant variation in baseline between the experimental group and comparators.

2.4. Data extraction

This research employed an adapted version of the Cochrane Good Practice Data Extraction form, ²⁷ and two researchers independently conducted the data extraction and entry. The data from included studies were extracted for the first author, publication period, eligibility criteria, participants (population, setting, sample size, age range, mean BMI, mean menstrual duration), intervention (the type of aerobic exercise, intensity, frequency, duration, period), comparators (type of comparator, frequency, and duration), and outcomes (indicators, measurement tool, variance) (Table 1). Authors would contact the writer via e‐mail for relevant data in cases of missing or unclear data.

TABLE 1.

Description of included studies.

Study	n	Participants	Age (years)	BMI (kg/m2)	Menstruation duration (day)	Interventions	Comparators	Intensity/duration/frequency/period	Outcome	Result
Arora et al., 2014 28	60	Young female students, India	19–24 years 20.37 ± 1.55	–	–	Race on a treadmill device	No treatment	High intensity 50 min, 3–5·week−1 3 menstrual cycles	Pain intensity via VAS	Pain intensity**
Hashemi et al., 2022 29	40	Students, Iran	18–26 years 22.15 ± 47.72	–	–	Water yoga exercises, under the supervision	Not to do any specific activity	45 min, 2·week−1 2 menstrual cycles	Pain intensity via MPQ Pain duration	Pain intensity** Pain duration**
Heidarimoghadam et al., 2019 30	86	Single students, Iran	18–24 years 18.65 ± 0.64	21.93 ± 3.02	5.74 ± 1.33	Aerobic exercises with the main walking base, under the supervision	2 pe classes 1 week−1	Low intensity 47 min, 2·week−1 2 Menstrual cycles	Pain intensity via MPQ Pain duration	Pain intensity** Pain duration**
Kannan et al., 2019 31	65	Women, New Zealand	18–43 years –	–	–	Race on a treadmill device under the supervision	No treatment	High intensity 50 min, 3·week−1 1 menstrual cycles at the school 6 Menstrual cycles at home	Pain intensity via VAS	Pain intensity**
Rakhshaee, 2011 32	92	Students, Iran	18–22 years –	–	–	Yoga, taught by a booklet	No treatment	20 min, 14 days of 4 week 3 Menstrual cycles	Pain intensity via VAS (0–3 scale) Pain duration	Pain intensity** Pain duration**
Reyhani, 2013 33	90	Nursing students, Iran	–	–	–	Brisk walking (1 training session)	Usual care	30 min, first 3 days of menstruation 3 menstrual cycles	Pain intensity via VAS	Pain intensity**
Samy et al., 2019 13	98	Women, Egypt	18–25 years 21.74 ± 1.47	20.97 ± 2.46	4.51 ± 0.94	Zumba is supervised by instructors and pamphlets	No treatment	60 min, 2·week−1 2 menstrual cycles	Pain intensity via VAS Pain duration	Pain intensity** Pain duration**
Siahpour et al., 2013 43	60	University students, Iran	20–25 years –	–	–	Yoga trained Aerobic dance	No exercise or yoga performed	60 min, 3·week−1 2 menstrual cycles	Pain intensity via VAS Pain duration	Pain intensity** Pain duration**
Song and Kim, 2023 15	30	Women, Korea	19–39 years 32.6 ± 1.41	21.30 ± 1.97	5.20 ± 1.23	Pilates sessions, led by one instructor	No treatment	50 min, 2·week−1 3 Menstrual cycles	Pain intensity via VAS	Pain intensity**
Temizkan and Budak, 2021 34	30	Young adults, Turkey	15–30 years 22.37 ± 2.14	21.03 ± 3.18	–	Walking and climbing stairs	No treatment	Moderate intensity 45 min, 3·week−1 3 menstrual cycles	Pain intensity via MPQ	Pain intensity**
Tharani et al., 2018 14	30	Students, India	17–23 years –	–	–	Aerobic dance, led by an instructor	Simple stretching	45 min, 3·week−1 2 menstrual cycles	Pain intensity via VAS Pain duration	Pain intensity** Pain duration**
Vaziri et al., 2014 35	70	University students, Iran	18–33 years 20.77 ± 1.97	24.44 ± 2.82	6.17 ± 1.26	Race on a treadmill device	No treatment	20 min, 3·week−1 3 Menstrual cycles	Pain intensity via MSQ	Pain intensity**
Yang and Kim, 2016 36	36	Nursing students, Korea	18–25 years –	20.65 ± 1.54	5.86 ± 2.09	Yoga under the guidance and supervision	Usual care Not to do yoga	60 min, 1·week−1 3 menstrual cycles	Pain intensity via VAS Pain duration	Pain intensity** Pain duration
Yilmaz‐Akyuz and Aydin‐Kartal., 2019 37	67	Freshman, Turkey	18–35 years 19.44 ± 1.92	21.39 ± 2.83	6.37 ± 1.77	Race on a treadmill device	No treatment	Low intensity 30 min, 3·week−1 3 Menstrual cycles	Pain intensity via VAS	Pain intensity**
Yonglitthipagon et al., 2017 38	34	University students, Thailand	18–22 years 19.83 ± 1.35	21.63 ± 3.63	3.80 ± 2.00	Yoga, taught by booklet	No treatment	30 min, 2·week−1 3 Menstrual cycles	Pain intensity via VAS	Pain intensity**
Lee and Shin, 2021 39	35	University students, Korea	20 years 22.09 ± 0.69	–	–	Race on a treadmill device Pilates, led by one instructor	No treatment	Moderate intensity (race) 50 min, 3·week−1 1 menstrual cycles	Pain intensity via VAS	Pain intensity** Pain intensity

Abbreviations: MPQ, Mc gill Pain Questionnaire; MSQ, Menstrual Symptom Questionnaire; VAS, Visual Analog Scale.

^** p < 0.01.

2.5. Quality assessment

2.5.1. Methodological quality

The two researchers used the Cochrane Collaboration risk of bias tool to assess the studies independently. Any disagreements were resolved through consultation with a third experienced researcher. Due to differences in exercise interventions, participants should be informed about the experiment's objective, process, and potential risks in advance. ⁴⁰ In addition, it's difficult to standardize or conceal operations and instructions that require active participation, making it challenging to implement the blind method. ²⁷ , ⁴¹ Thus, despite their high risk of bias, studies meeting these conditions were still considered high quality and included in the study. ¹² , ⁴²

2.5.2. Quality of evidence

This study used the online GRADE pro‐GDT tool to evaluate the quality of evidence concerning the risk of bias, inconsistency, indirectness, precision, and publication bias. The quality of evidence was categorized as “high,” “moderate,” “low,” and “very low.” To ensure the reliability and applicability of evidence, two researchers independently conduct assessments, with any disagreements resolved together with a third researcher.

2.6. Data synthesis

Review Manager 5.4 and Stata 17 were used for systematic review and meta‐analyses of pain intensity and pain duration, according to the Cochrane review protocol. Pain intensity, measured by validated tools, was reported using standard mean difference (SMD) and 95% CI. Pain duration was measured in hours and reported using weighted mean difference (WMD) and 95% CI. Heterogeneity was evaluated using Cochrane Q statistics. ²⁵ When p < 0.01, I ² >50% indicated notable heterogeneity. A random‐effect model would be employed to combine the effect size as anticipated there may be a high degree of heterogeneity. Subgroup analysis of aerobic exercise components (type, intensity duration, frequency, and period) would be made to investigate the dose response relationships. Trials comparing two aerobic exercise interventions against one comparator were considered two separate trials. ² , ²⁷ The number of participants in comparators would be divided averagely to the experiments to avoid repeated calculation. ²⁸ The stability of the combined effect was assessed by the one‐by‐one exclusion method. Publication bias was tested using the Egger test and funnel plot.

3. RESULTS

A total of 1351 articles were identified by searching various databases, citations, and related websites. Two researchers carefully screened the articles by removing duplicates, reviewing titles and abstracts, assessing inclusion and exclusion criteria, and evaluating experiment design and the statistical indicators of outcomes. Ultimately, 18 RCTs from 16 articles were included. Detailed reasons for exclusion and the screening process are presented in Figure 1.

FIGURE 1.

PRISMA flow diagram.

3.1. Study characteristics

This research identified 18 RCTs from 16 articles, with 15 of the studies involving adolescents. These studies were mainly published between 2011 and 2023, involving 918 participants aged 15 to 43, with an average age of 21.26 ± 13.15. ¹³ , ¹⁴ , ¹⁵ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴³ The average BMI was 21.80% ± 2.98%, and the average duration of menstruation was 5.43 ± 1.67 days. Aerobic exercise interventions included running, yoga, dance, Zumba, FIIT training, Pilates, and brisk walking. The comparators included usual care, simple stretching, PE classes, and no treatment. The intensity was involved in low, moderate, and high, and each session lasted 20 to 60 min. The frequency can range from 1 to 5 times per week, and the period varies from one to three menstrual cycles. All RCTs used pain intensity as the outcome; the Visual Analog Scale (VAS), Menstrual Symptoms Questionnaire (MSQ), and the McGill Pain Questionnaire (MPQ) were the main measurement methods. Seven RCTs used pain duration as the outcome. All RCTs reported a significant effect in alleviating PD, except for Yang and Kim's study on pain intensity ³⁶ and Lee's study on pain duration. ³⁹ These RCTs required participants to exercise throughout the menstrual cycle except Rakhshaee and Reyhani. Rakhshaee et al. required yoga during the luteal phase of the second and third cycles, ³² while Reyhani et al. required brisk walking during the first 3 days of the cycle. ³³ Four studies reported dropout numbers, ¹³ , ²⁹ , ³¹ , ³⁴ among which Vaziri's study did not mention whether the dropouts belonged to the experiment or comparator groups. ³⁵ No studies mentioned adverse events. (Table 2).

TABLE 2.

Subgroup analysis results.

Subgroup	N	N (Ex/Co)	Q	Hedges		SMD	CI 95%	p
				H 2	I 2/%
Overall	18	459/429	142.63	9.38	89.34	−1.78	−2.26 to −1.31	0.00
Type of aerobic exercise
Race	5	143/136	35.04	1.00	85.96	−1.62	−2.37 to −0.87	0.00
Yoga	5	125/107	16.64	5.05	80.18	−1.68	−2.38 to −0.97	0.00
Dance	3	84/74	39.75	18.46	94.58	−2.24	−4.11 to −0.36	0.00
Walking	3	103/103	5.61	2.72	63.23	−0.98	−1.47 to −0.48	0.06
Pilates	2	27/21	0.54	1.00	0.00	−3.30	−4.17 to −2.43	0.46
	18	482/441	142.63	9.38	89.34	−1.78	−2.26 to −1.31	0.00
Intensity
High	2	62/63	18.81	18.81	94.68	−1.19	−2.89 to 0.52	0.00
Moderate	2	26/21	1.34	1.34	25.64	−0.96	−1.68 to 0.25	0.25
Low	2	78/75	1.55	1.55	35.56	−1.65	−2.10 to −1.19	0.21
	6	166/159	28.88	5.31	81.15	−1.30	−1.88 to −0.73	0.00
Duration, min
≤30	5	182/171	20.49	5.05	80.21	−1.33	−1.86 to −0.81	0.00
31–45	3	50/50	13.54	7.03	85.77	−1.45	−2.62 to −0.27	0.00
46–60	10	250/220	116.10	14.33	93.02	−2.06	−2.94 to −1.19	0.00
	18	482/441	156.37	11.88	91.58	−1.74	−2.27 to −1.21	0.00
Frequency, week−1
≤2	6	162/162	65.56	13.28	92.47	−2.23	−3.28 to −1.18	0.00
=3	9	195/162	48.98	5.96	82.93	−1.61	−2.21 to −1.00	0.00
<3	2	80/72	2.34	2.34	57.22	−1.73	−2.32 to −1.14	0.13
	17	437/396	126.89	8.66	88.45	−1.85	−2.34 to −1.36	0.00
Period, menstrual cycle
≤1	4	70/60	19.36	11.17	91.05	−1.43	−2.83 to −0.03	0.00
=2	6	167/147	51.29	9.17	89.70	−2.21	−3.13 to −1.28	0.00
≥3	8	245/234	39.46	6.10	83.60	−1.66	−2.18 to −1.13	0.00
	18	482/441	142.63	9.38	89.34	−1.78	−2.26 to −1.31	0.00

3.2. Risk of Bias of Included Studies

Ten of the sixteen studies explicitly reported the method of generating randomized sequences were evaluated as low risk, and four of these studies properly described the method of allocation concealment. ¹³ , ³¹ , ³⁷ , ³⁸ All blinding methods were rated as high risk due to the need for informed consent, self‐reported assessments, and variability in exercise interventions. ⁴⁰ , ⁴¹ Two studies may be biased due to data completeness, while others were evaluated as low risk. Vaziri et al. did not specify the groups for lost visits and reasons. ³⁵ Tharani did not report dropouts at the follow‐up. ¹⁴ Four of these studies provided study protocol or prespecified registration information, ¹³ , ²⁹ , ³¹ , ³⁴ rated as low risk, while the other literature lacked sufficient evidence. No other bias issues were reported (Figure 2).

FIGURE 2.

Distribution of methodological quality of included studies.

3.3. Overall effects of aerobic exercise on PD

3.3.1. Pain duration effects of aerobic exercise on PD

Seven studies reported the effect of aerobic exercise on alleviating pain duration in PD. Effect sizes were combined using the WMD scale. The random effect model was used due to high heterogeneity (I ² = 90.11%, p = 0.00). A “−” before the effect size indicated that aerobic exercise could shorten the pain duration of PD. Figure 3 shows that aerobic exercise could shorten the pain duration of PD (WMD of −12.53 h, p = 0.01, 95% CI [−21.38 to −3.68]).

FIGURE 3.

Effect size forest plot for pain duration.

3.3.2. Pain intensity effects of aerobic exercise on PD

Eighteen studies reported the effect of aerobic exercise on alleviating pain intensity in PD. Effect sizes were combined using SMD as scale. The random effect model was used due to high heterogeneity (I ² = 89.34%, p = 0.00). A “‐” before the effect size indicated that aerobic exercise could alleviate the pain intensity of PD. Figure 4 shows that aerobic exercise could alleviate the pain intensity of PD (SMD = −1.728, p = 0.00, 95% CI [−2.26 to −1.31]).

FIGURE 4.

Effect size forest plot for pain intensity.

3.4. Subgroup analysis of aerobic exercise variables

The overall effect revealed that although aerobic exercise could effectively reduce pain intensity in PD, the combined effect sizes were highly heterogeneous (I ² = 88%, p < 0.01), suggesting that moderating variables might influence the effect. Considering this, this study follows the subgroup analysis from the aspects including the type of aerobic exercise (running, yoga, dance, brisk walking, Pilates), intensity (Low, Moderate, High), duration (≤30 min, 31–45 min, 47–60 min), frequency (≤two times per week, three times per week, >three times per week), and menstrual cycle (≤one menstrual cycle, two menstrual cycles, ≥three menstrual cycles). The criteria for the category of exercise intensity were based on the American College of Sports Medicine's Exercise Intensity Classification. ⁴⁴

3.4.1. Type

The 18 RCTs were categorized into five groups: running ²⁸ , ³¹ , ³⁵ , ³⁷ , ³⁹ (five studies), yoga ²⁹ , ³² , ³⁶ , ³⁸ , ⁴³ (five studies), dance (three studies), ¹³ , ¹⁴ , ⁴³ walking ³⁰ , ³³ , ³⁴ (three studies), and Pilates (two studies). ¹⁵ , ³⁹ Among them, the Pilates group showed the maximum effect size in PD's pain intensity alleviation (SMD = −3.30, 95% CI [−4.17 to −2.43]), followed by the dance (SMD = −2.24, 95% CI [−4.11 to −0.36]), yoga (SMD = −1.68, 95% CI [−2.38 to −0.97]), running (SMD = −1.62, 95% CI [−2.37 to −0.87]), and walking (SMD = −0.98, 95% CI [−1.47 to −0.48]).

3.4.2. Intensity

The six studies were categorized into three groups: low ³⁰ , ³⁷ (two studies), moderate (two studies), ³⁴ , ³⁹ and high ²⁸ , ³¹ (two studies). Among them, the low‐intensity group showed the maximum effect size in PD's pain intensity alleviation (SMD = −1.65, 95% CI [−2.10 to −1.19]), followed by high‐intensity group (SMD = −1.19, 95% CI [−2.89 to −0.52]), and moderate‐intensity group (SMD = −0.96, 95% CI [−1.68 to −0.25]).

3.4.3. Duration

The 16 studies were categorized into three groups: ≤30 min (four studies), ³² , ³³ , ³⁵ , ³⁷ , ³⁸ 31–45 min ¹⁴ , ²⁹ , ³⁴ (five studies), and 46–60 min ¹³ , ¹⁵ , ²⁸ , ³⁰ , ³¹ , ³⁶ , ³⁹ , ⁴³ (eight studies). Among them, the 46–60 min group showed the maximum effect size in PD's pain intensity alleviation (SMD = −2.06, 95% CI [−2.94 to −1.19]), followed by 31–45 min group (SMD = −1.45, 95% CI [−2.62 to −0.27]), and ≤30 min group (SMD = −1.33, 95% CI [−1.86 to −0.81]).

3.4.4. Frequency

The 17 studies were categorized into three groups: ≤two times per week ¹³ , ¹⁵ , ²⁹ , ³⁰ , ³⁶ , ³⁸ (six studies), three times per week (nine studies), ¹⁴ , ³¹ , ³⁴ , ³⁵ , ³⁷ , ³⁹ , ⁴³ and >three times per week (two studies). ²⁸ , ³² The ≤two times per week group showed the maximum effect size in PD's pain intensity alleviation (SMD = −2.23, 95% CI [−3.28 to −1.18]), followed by the >three times per week group, (SMD = −1.73, 95% CI [−2.32 to −1.14]), and three times per week group (SMD = −1.61, 95% CI [−2.21 to −1.00]).

3.4.5. Period

The 17 studies were categorized into three groups: ≤one menstrual cycle ³¹ , ³⁴ , ³⁹ (two studies), two menstrual cycles ¹³ , ¹⁴ , ²⁹ , ³⁰ , ⁴³ (six studies), and ≥three menstrual cycles ¹⁵ , ²⁸ , ³² , ³³ , ³⁵ , ³⁶ , ³⁷ , ³⁸ (eight studies). The two Menstrual cycle group showed the maximum effect size in PD's pain intensity alleviation (SMD = −2.21, 95% CI [−3.13 to −1.28]), followed by the ≥three Menstrual cycle group (SMD = −1.66, 95% CI [−2.18 to −1.13]), and ≤1 Menstrual cycle (SMD = −1.43, 95% CI [−2.83 to −0.03]).

3.5. Sensitive analysis

Stata 17 was used for sensitivity analysis and indicated that the meta‐results were rather stable; while any literature was excluded, the overall results did not change significantly (Figure 5).

FIGURE 5.

Sensitivity analysis plot.

3.6. Publication bias

Visual inspection of the funnel plot of the literature related to pain intensity showed a slight asymmetry (Figure 6), which, together with Egger's regression test outcomes (b = −2.18, p = 0.045 < 0.01), suggests the potential of publication bias. ⁴⁵ No publication bias assessment was conducted for pain duration literature due to the limited number of studies.

FIGURE 6.

The funnel plot of publishing bias for pain intensity.

3.7. Quality of evidence

The overall evaluation of evidence strength using GRADE indicated moderate quality for pain intensity and low quality for pain duration (Figure 7). This was mainly due to performance bias, detection bias, and small sample size (only for pain duration). Therefore, these findings should be explained cautiously. However, excluding the low‐quality studies, the pain intensity and duration results remained stable, providing some reassurance about the observed treatment effect.

FIGURE 7.

GRADE evidence profile for pain intensity and pain duration.

4. DISCUSSION

Pain duration is a crucial parameter to describe the pain and intervention process. ⁴⁶ Effective interventions can shorten pain duration, which plays a vital role in alleviating individual suffering, preventing drug dependence, enhancing treatment adherence, and minimizing economic burdens. ⁶ , ¹¹ , ⁴⁷ The result of the Meta‐analysis demonstrated that aerobic exercise could effectively alleviate pain duration (WMD = −12.53 h, p = 0.01, 95% CI: −21.38 to −3.68, GRADE = Low). This aligns with studies by Matthewman and Carroquino‐Garcia. ²⁴ , ²⁷ Aerobic exercise can accelerate pain alleviation by stimulating PGs' synthesis of inhibitors, increasing the rate of PG's metabolism, and aiding in eliminating acidic metabolites. ¹⁶ Specifically, during menstruation, there is an increase in the secretion of prostaglandins in the uterus, which leads to accelerated uterine contractions and induces ischemia and hypoxia in myometrial tissues. ¹⁸ At the same time, The metabolism of PG's rate began to decrease, and acidic metabolites piled up in the oxygen‐free environment, ¹⁸ ultimately prolonging the pain. Aerobic exercise can activate the uterine PG synthetic inhibitor to relieve uterine spasms and enhance blood circulation and oxygen supply to raise the rate of PG's supersession, fully oxidize the acidic metabolites, ⁴⁸ and expedite their excretion to accelerate pain relief. However, these results should be explained carefully due to the clinical heterogenicity of the limited number of studies and individual variations in pain perception. In subsequent research, employing non‐self‐reported and objective pain assessment metrics should be considered to enhance result applicability.

Pain intensity serves as the most straightforward standard for assessing pain management efficacy, and achieving effective pain relief is the primary aim of aerobic exercise interventions for PD. ¹² , ⁴⁹ This is especially important for adolescents and young women to maintain their work or learning performance and quality of life during menstruation. ⁶ , ³¹ , ³⁸ Meta‐analysis results showed that aerobic exercise effectively alleviated pain intensity, compared to the comparator group (SDM = −1.72, p = 0.00, 95% CI: −2.21 to −1.22, GRADE = Moderate). This conclusion aligns with the majority of meta‐analyses on this topic. ² , ²³ , ²⁴ , ²⁷ It has been shown that aerobic exercise can alleviate pain intensity by regulating hormonal changes in the endometrium and increasing the pain threshold by stimulating the release of multiple analgesic drugs, acting on inflammatory mediators, and mobilizing positive emotions and other mechanisms. ⁴ , ¹⁷ , ⁴⁹ Specifically, monthly repeated pain of women with PD may heighten their pain sensitivity of the nerve center, ¹⁹ , ⁴⁸ lower pain tolerance, and may even trigger emotions such as depression and anxiety. ⁷ , ⁵⁰ Aerobic exercise can stimulate the release of pain‐relieving drugs like endorphins, cannabinoids, and anti‐inflammatory macrophages, reduce pain sensitivity, and inhibit proinflammatory factors' activity. ¹⁶ , ¹⁸ Thus, it can effectively raise the pain threshold and relieve inflammation and swelling ¹⁰ so that the pain will be relieved. It also triggers the release of neurotransmitters like dopamine and serotonin, which help promote pleasure and relieve depression. ⁷ , ⁵¹

While aerobic exercise is acknowledged as the preferred complementary therapy for pain management, ²⁴ no formal statistical comparisons of aerobic exercise type or dose were made, limiting the conclusions that can be drawn. ⁵² , ⁵³ This study aimed to explore heterogeneity and evaluate dose–response relationships by investigating the potential effect of aerobic exercise dose components (intensity, duration, frequency, and period). The objective is to provide insights into the development of tailored exercise prescriptions for adolescents and young women experiencing PD.

The type test indicated that the Pilates group yielded the maximum effect, followed by aerobic dance, yoga, running, and brisk walking. This finding is consistent with the results of Luana's study, which confirmed a significant reduction in pain intensity after Pilates exercise. ⁵¹ Previous research indicated that Pilates movements, especially those focusing on the pelvic region, play a positive role in blood circulation in the pelvic area and organ function. ⁵⁴ Compared to other aerobic exercises, this can offer more specific relief for uterine spasm pain caused by ischemia and hypoxia, ³⁹ leading to a more significant reduction in pain.

The test on aerobic exercise dose components revealed that the optimal effect was achieved with low intensity, 46–60 min, two menstrual cycles, and ≤two times per week in their respective groups, and it demonstrates a dose–response (Figure 8).

FIGURE 8.

Dose–response effects between elements of aerobic exercise and pain intensity.

The intensity test indicated that different levels of aerobic exercise can help alleviate the pain intensity of PD, with the low‐intensity group showing the most significant improvement. However, it differed from previous studies' results, which found that moderate‐ or high‐intensity aerobic exercises were notably superior to low‐intensity exercises in alleviating PD pain. ²⁵ Additionally, Kelli's study observed that high‐intensity exercise was more accessible to arouse Exercise‐induced Hypoalgesia from the perspective of endogenous pain modulation, which was also different from this study. ⁵⁵ , ⁵⁶ This could be attributed to the participants' higher adherence to low‐intensity exercise. ⁵⁷ Furthermore, in Heidarimoghadam's study, the low‐intensity group's experimental exercise process was strictly monitored, to some extent influencing the meta‐analysis outcomes. ³⁰ , ⁴⁴

The duration test results indicated that exercise lasting between 20 to 60 min all can alleviate the pain intensity of PD, with the most effective duration being 46 to 60 min. There was a positive correlation between the exercise duration and pain‐relieving effect, suggesting that more extended exercise within this range increased pain relief. However, many adolescents and women reduce or skip exercise during their periods. ⁴ , ⁵ , ⁵⁸ Even exercise below this range is still helpful for those incapable or unwilling to achieve the amount of exercise. ⁵² The intensity and volume of exercise should be adjusted according to the change in the female body during PD. ⁵²

The menstrual cycle test indicated that one to three menstrual cycle exercises all alleviate the pain intensity of PD, with the two menstrual cycle group being the most effective. As the period increases, the effect first increases and then decreases. Physiological studies have shown that aerobic exercise stimulates the release of endogenous opioids, ⁵⁰ and concentrations gradually increase with 5–8 weeks of continuous exercise. However, long‐term systematic exercise leads to a compensatory reduction of opioids and their receptors. ⁴⁸ The finding aligns with Oliveira's study, which confirmed the impact of aerobic exercise on the expression of opioid receptors in the hippocampal structure of rats. ¹⁶ That study indicated that the expression effect started diminishing once the training was more than 9 weeks or 45 repetitions. ¹⁶ , ¹⁹

The frequency test found that one to three times per week all effectively alleviated the pain intensity of PD, with the three times per week group showing the most significant result. Some studies indicate that women who exercise regularly have a lower incidence of PD compared to those who exercise occasionally. ⁸ , ¹⁰ However, some studies suggest that pain alleviation is more directly related to exercise intensity and volume than frequency. ¹⁶ , ⁵⁹ Some studies even showed that excessive exercise frequency might lead to menstrual dysfunction. ⁵⁹ Since this study could not effectively control factors like age, number of births, smoking, stress, mood, oral contraceptive pill use, and the effect of the blinding method. The relationship between the volume of exercise and the pain could not be reflected accurately. More high‐quality RCTs are needed in the future to confirm a causal biological effect.

In conclusion, the “some is good, more is better” principle applies to exercise dosage. ⁶⁰ However, considering the interactions between exercise dose components and individual response variability, ⁵² , ⁶¹ extra and targeted RCTs and meta‐analyses with large sample sizes are still needed in the future to explore the interactions between populations of different fitness levels and the components of aerobic exercise dosage.

This study conducted a systematic review strictly according to the Cochrane Handbook and reported following the PRISMA statement. A comprehensive search of databases and scientific research websites about women and dysmenorrhea without language restrictions, dates, publishing types, or RCT filters. Articles of different languages were included to minimize language bias. Moreover, this study is the first to investigate the relationship between pain relief and dose components of aerobic exercise, providing medical evidence for subsequent exercise prescription and pain management of PD.

This study provides some valuable insights but still has limitations. The quality of the included literature could have been higher due to the performance bias and selective outcome reporting. Despite efforts to identify gray literature, publication bias still existed and should be interpreted carefully, combined with clinical knowledge. ¹ , ²⁷ , ⁴⁹ Both outcomes showed high heterogeneity, only partially alleviated by subgroup analyses of pain intensity, but were still influenced by factors like sample selection, recall bias, poor blinding implementation, and social expectations. ¹² , ²⁷ Few studies had intervention cycles longer than three menstrual cycles, limiting the assessment of effects for long‐term or stopping exercising Furthermore, there have been no RCTs that have specifically focused on adolescents under 18, or directly compared the alleviation of aerobic exercise for PD between adolescents and young women. This lack of research has hindered age‐specific analyses and recommendations.

5. CONCLUSION

This systematic review and meta‐analysis showed that aerobic exercise was able to alleviate pain intensity and duration in adolescents and young women with PD. Moderate‐quality evidence suggests a dose‐effect relationship between exercise dose and the alleviation of PD pain intensity, and low intensity, 46–60 min, ≤2 times per week, and two menstrual cycles showed more efficiency in alleviating PD.

Future studies should encompass diverse age groups, special populations, longer intervention cycles, and placebo designs. The priority should be RCTs of aerobic exercise interventions in younger age groups. These trials should follow reporting standards, use non‐self‐report outcomes, and detail exercise dosage to develop personalized adolescent treatment strategies. Additionally, a better understanding of how aerobic exercise enhances the quality of life in women with PD is required. Furthermore, these results should be interpreted within a clinical context.

AUTHOR CONTRIBUTIONS

Jingjie Cai, Mingyi Liu, and Yan Jing substantially contributed to the conception and design. Zikang Yin and Nianxin Kong contributed to data acquisition, analysis, and interpretation. Jingjie Cai and Yan Jing drafted the manuscript for important content. Jingjie Cai and Mingyi Liu critically revised the manuscript for important intellectual content. All authors gave their final approval of the manuscript for publication.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Supporting information

Appendix S1.

Cai J, Liu M, Jing Y, Yin Z, Kong N, Guo C. Aerobic exercise to alleviate primary dysmenorrhea in adolescents and young women: A systematic review and meta‐analysis of randomized controlled trials. Acta Obstet Gynecol Scand. 2025;104:815‐828. doi: 10.1111/aogs.15042