Received: 10 May 2022 | Revised: 13 August 2022 | Accepted: 19 August 2022

DOI: 10.1111/aogs.14449

S Y S T E M AT I C R E V I E W

The relation between body mass index and primary

dysmenorrhea: A systematic review and meta-­analysis

Lingsha Wu | Jing Zhang | Jie Tang | Haiyan Fang

Nursing School, Anhui University of

Chinese Medicine, Hefei, Anhui, China Abstract
Introduction: The relation between body mass index (BMI) categories and the occur-
Correspondence
Haiyan Fang, Nursing School, Anhui rence of primary dysmenorrhea has been investigated, but the results of these studies
University of Chinese Medicine, Hefei,
are inconsistent and controversial. The aim of our study was to systematically review
Anhui, China.
Email: 1269807949@qq.com the literature and investigate the association between each category of BMI and the
occurrence of primary dysmenorrhea.
Funding information
Key Project of Natural Science Research Material and methods: We conducted a systematic review and meta-­analysis of ob-
Fund for Universities of Anhui Province
servational studies related to BMI and primary dysmenorrhea. Eleven databases—­
Education Department, Grant/Award
Number: KJ2020A0444 PubMed, Medline, Embase, Cumulative Index to Nursing and Allied Health Literature
(CINAHL), Scopus, Cochrane Library, Web of Science, China National Knowledge
Infrastructure (CNKI), Chinese Science and Technology Periodical Database (VIP),
Chinese Biomedical Literature Database (CBM), and Wanfang database—­were sys-
tematically searched from inception to March 2022. We used the 11 items recom-
mended by the Agency for Healthcare Research and Quality to assess the quality of
included studies. The Q test and the I2 test were used to evaluate the heterogeneity
among studies. Odds ratios (OR) and their 95% confidence intervals (CI) were pooled
by fixed-­effects models or random-­effects models. Stata software version 16.0 was
used to complete the statistical analyses.
Results: A total of 4181 articles were collected from the database, and 12 studies
were included based on inclusion and exclusion criteria. A total of 29 647 participants
were included in the study, with a mean baseline age of 17–­45 years. All included lit-
erature was published between 2017 and 2021 and was conducted in six countries.
Eleven included studies were of medium quality and one included study was of high
quality. Being underweight may be related to the occurrence of primary dysmenor-
rhea (12 studies, n = 6545, OR 1.43, 95% CI 1.18–­1.73). Being overweight (12 studies,
n = 3098) and obesity (four studies, n = 94) may not be associated with the develop-
ment of primary dysmenorrhea.
Conclusions: Being underweight may increase the risk of the occurrence of pri-
mary dysmenorrhea, whereas overweight and obesity might not be associated with

Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction
in any medium, provided the original work is properly cited and is not used for commercial purposes.
© 2022 The Authors. Acta Obstetricia et Gynecologica Scandinavica published by John Wiley & Sons Ltd on behalf of Nordic Federation of Societies of
Obstetrics and Gynecology (NFOG).

1364 | 
wileyonlinelibrary.com/journal/aogs Acta Obstet Gynecol Scand. 2022;101:1364–1373.
WU et al. | 1365

primary dysmenorrhea. Due to the limitations of the meta-­analysis, more stud-

ies are needed to investigate the relation between each category of BMI and the
occurrence of primary dysmenorrhea. To maintain a balanced diet and an appro-
priate lifestyle is beneficial for people to have the normal category of BMI and
live a healthy life, which may play a role in preventing the occurrence of primary
dysmenorrhea.

KEYWORDS
body mass index, meta-­analysis, primary dysmenorrhea, relationship

1 | I NTRO D U C TI O N

Primary dysmenorrhea is defined as periodic abdominal cramps oc- Key message

curring before or during menstruation, in the absence of other or- This review suggests that being underweight may increase
ganic diseases.1 The incidence of primary dysmenorrhea is high and the risk of the occurrence of primary dysmenorrhea,
2
shows an increasing trend year by year. Primary dysmenorrhea se- whereas being overweight and obese may not be associ-
riously affects women's daily activities and quality of life. Not only ated with the primary dysmenorrhea. More high-­quality
does it cause physical discomfort, such as headache, vomiting, fa- studies in these areas are needed in the future.
tigue, and back pain, but also it can negatively affect mental health,
thus further aggravating the relevant symptoms, forming a vicious
circle.3–­5 Related research has shown that family history of primary
dysmenorrhea, menstrual cycle length, dietary habits, and body 2.2 | Data sources: search strategy and
mass index (BMI) are the related factors in the occurrence of primary selection criteria
dysmenorrhea.6,7 However, our knowledge of these connections is
limited. We searched literature published in PubMed, Medline, Embase,
Body mass index, an anthropometric assessment measure de- Cumulative Index to Nursing and Allied Health Literature (CINAHL),
fined as body weight measured in kilograms divided by the square Scopus, Cochrane Library, Web of Science, China National
of height in meters, is a valuable measure of the nutritional sta- Knowledge Infrastructure (CNKI), Chinese Science and Technology
tus of a person. 8,9 The World Health Organization classifies BMI Periodical Database (VIP), Chinese Biomedical Literature Database
into four categories: underweight, normal weight, overweight, (CBM), and the Wanfang database from their inception through
and obesity.10 Although the cut-­off values of BMI are different in March 2022, without language restrictions. Meanwhile, we manu-
various parts of the world, they are all grouped into the four cat- ally retrieved the list of references for each included study to iden-
egories mentioned above. Some studies have revealed that BMI tify other potentially relevant articles. The medical subject words
was associated with the occurrence of primary dysmenorrhea.11–­13 and keywords used in the search were: “BMI”, “body mass index”,
Some researchers further pointed out that a lower or higher BMI “obesity”, “weight”, “dysmenorrhea”, “pain,menstrual”, “primary
14
increases the risk of the occurrence of primary dysmenorrhea, dysmenorrhea”, “functional dysmenorrhea”, “menorrhalgia”, “men-
whereas other researchers believe that there was no significant strual cramps”. Studies were included if (a) the participants were
connection between them.15,16 women with primary dysmenorrhea, (b) the studies examined the
Although several studies have explored the relation between relation between BMI and primary dysmenorrhea, (c) the designs
BMI and the occurrence of primary dysmenorrhea, these findings were population-­based observational studies, including cross-­
are inconsistent and the results are controversial. This meta-­analysis sectional studies, case–­control studies and cohort studies, and (d)
aimed to synthesize the association between each category of BMI the studies included three or more categories for BMI, because we
and the occurrence of primary dysmenorrhea. wished to perform a detailed study of the relation between under-
weight, overweight, and obese BMI and the occurrence of primary
dysmenorrhea. We excluded studies that were reviews, case re-
2 | M ATE R I A L A N D M E TH O DS ports, comments, conference papers, animal experiments, and un-
published studies; duplicate publications (keeping the study with
2.1 | Protocol and registration the most information only); and literature with incomplete data.
Screening of studies was performed independently by two authors
This systematic review was conducted following PRISMA recom- (LW, JT), and disagreements were resolved through discussion (LW,
mendations,17 with registration no. CRD42022324984. JT, HF).
1366 | WU et al.

2.3 | Outcome measure up the scores of the 11 items. The quality of the articles was assessed
as follows: a total score of 0–­3 was considered low quality, 4–­7 was
The primary outcome was the odds ratio (OR) and its 95% confi- considered medium quality, and 8–­11 was considered high quality.19
dence interval (CI) indicating the relation between each category of
BMI and the occurrence of primary dysmenorrhea.
2.5 | Statistical analyses

2.4 | Data extraction and quality assessment The effect sizes of the relation between the categories of BMI and pri-
mary dysmenorrhea were expressed as OR with 95% CI. In detail, we used
The data extraction was performed by two authors (LW, JZ) indepen- the normal category of BMI in each included study as a reference group
dently. Differences of opinion among the reviewers were resolved in (OR 1) and then compared it with underweight (<18.5 kg/m2), overweight
extensive discussion (LW, JZ, HF). As a result, consensus was reached (25–­29.9 kg/m2) and obese (≥30 kg/m2) BMI, respectively. The heterogene-
on the inclusion of the following information: the name of the first au- ity among studies was assessed using the Q test and the I2 test.20 According
thor, publication year, country, study design, study period, sample size, to the calculation results, we used a fixed-­effects model when heterogene-
age of women with primary dysmenorrhea, continent, pain measure, ity was not statistically different (p > 0.05 or I2 < 50%); on the contrast, a
BMI categories, and the corresponding OR and their 95% CI. Three au- random-­effects model was used. Then, we performed subgroup analyses
thors (LW, JZ, JT) independently used the 11 items recommended by of the results with high heterogeneity to detect its sources,21 with the fol-
Agency for Healthcare Research and Quality18 to evaluate the quality lowing prespecified subgroups: continents (Asia, Europe). Begg's tests and
of cross-­sectional studies. Differences were resolved by careful com- Egger's tests were performed to evaluate the risk of bias.22 The results of
munication (LW, JZ, JT, HF). All 11 entries were evaluated with “yes”, all meta-­analyses were presented as the relevant OR and 95% CI, I2 and
“no”, and “unclear” and scored as 1 for “yes” and 0 for “no” or “unclear”. p values. In this study, all statistical analyses were performed using Stata
The total score, which ranged from 0 to 11, was obtained by adding software version 16.0 (StataCorp).

Records identified Additional records

through database identified through
searching (n = 4155) citation tracing (n = 26)

Records after duplicates removed

(n = 2537)

Records excluded after screening of

title and abstract (n = 2300)

Full-text studies assessed for

eligibility (n = 237)

Full-text studies excluded with

reasons (n = 225)
Non-primary dysmenorrhea of
women (n = 35)
Not exploring the relation
between BMI and PD (n = 87)
No population-based observational
studies (n = 7)
No or incomplete data (n = 96)

Studies included in the meta-analysis

(n = 12) F I G U R E 1 Selection process for
literature inclusion in the meta-­analysis
WU et al. | 1367

TA B L E 1 Basic information of the included literature

Sample size
Author/year/ (n)/ age (years)/ BMI
countryref Study design Study period continent Pain measure categories OR (95% CI)
a
Wang/2017/ Cross-­sectional Not reported 4630/17–­22/Asia Guiding principles for <18.5 1.781 (1.568–­2.023)
China23 clinical research 18.5–­23.9b 1
of new Chinese ≥24c 0.683 (0.574–­0.813)
medicine
Jing/ Cross-­sectional October–­ 961/ Guiding principles for <18.5a 1.235 (0.910–­1.675)
2018/ December, 17–­25/ clinical research 18.5–­24.9b 1
China24 2017 Asia of new Chinese ≥25c 0.777 (0.377–­1.603)
medicine
Zurawiecka/2018/ Cross-­sectional 2015–­2016 771/19–­25/ The Andersch and <18.5a 3.280 (1.760–­6.110)
Poland25 Europe Milsom scale 18.5–­25b 1
>25c 2.310 (1.230–­4.330)
Rafique/ Cross-­sectional March, 2016–­ 370/ The numeric pain <18.5a 2.401 (1.076–­5.358)
2018/ March, 18–­25/ relating scale 18.5–­24.99b 1
Arabia26 2017 Asia 25–­29.99c 1.957 (0.961–­3.985)
≥30 d 5.335 (1.215–­23.416)
Fernández-­ Cross-­sectional May–­June, 258/ The visual analog <18.5a 1.457 (0.565–­3.757)
Martínez/ 2017 18–­45/ scale 18.5–­24.99b 1
2018/ Europe ≥25c 0.971 (0.425–­2.221)
Spain27
Wang/ Cross-­sectional March–­July, 1069/ Obstetrics and ≤18.4a 0.786 (0.577–­1.072)
2019/ 2018 18–­25/ gynecology 18.5–­23.9b 1
China28 Asia 24–­27.9c 0.659 (0.357–­1.217)
≥28d 0.844 (0.216–­3.293)
Zheng/ Cross-­sectional Not reported 1200/ The visual analog <18.5a 2.032 (1.496–­2.761)
2020/ 15.80 ± 2.80/ scale 18.5–­23b 1
China29 Asia >23c 1.080 (0.816–­1.428)
Jiang/ Cross-­sectional Not reported 14 828/ Gynecology of <18.5a 1.216 (1.073–­1.379)
2020/ 18–­45/ Traditional 18.5–­23.9b 1
China30 Asia Chinese Medicine ≥24c 1.046 (0.901–­1.215)
Hu/ Cross-­sectional September, 4428/ The visual analog <18.5a 1.249 (1.090–­1.431)
2020/ 2017–­June, 19.00 ± 1.20/ scale 18.5–­24b 1
China31 2018 Asia ≥24c 0.965 (0.733–­1.272)
Hashim/ Cross-­sectional September, 336/ Not reported <18.5a 1.060 (0.450–­2.450)
2020/ 2017–­May, 19–­26/ 18.5–­24.9b 1
Arabia7 2018 Asia 25–­29.9c 1.140 (0.550–­2.380)
≥30 d 0.970 (0.340–­2.750)
Shellasih/ Cross-­sectional July 17–­24, 246/ Not reported <18.5a 0.940 (0.454–­1.949)
2020/ 2018 not reported/ 18.5–­24.9b 1
Indonesia32 Asia ≥25c 1.011 (0.444–­2.304)
Karout/ Cross-­sectional April–­July, 550/ The visual analog <18.5a 2.013 (0.882–­4.594)
2021/ 2019 18–­3 0/ scale 18.5–­24.9b 1
Lebanon33 Asia 25–­29.9c 1.650 (0.835–­3.263)
≥30 d 1.445 (0.411–­5.077)

Note: aunderweight; bnormal weight; coverweight; dobesity.

Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio.

3 | R E S U LT S studies. Twelve studies were included7,23–­33 in this meta-­analysis

(Figure 1). Table 1 shows the basic information of the included
3.1 | General study characteristics literature. The 12 included studies were all cross-­s ectional stud-
ies, with a total sample size of 29 647 participants, with a mean
A total of 4181 articles were collected from databases (4155 arti- baseline age of 17–­45 years. All studies were published between
cles from searched databases and 26 additional articles obtained 2017 and 2021. The study period ranged from 1 week to 1 year
through citation tracing); 2300 studies were excluded after read- for the included studies, except for three studies23,29,30 that did
ing titles and abstracts, followed by reading the full text of 237 not report the study period. All included studies were conducted
1368 | WU et al.

TA B L E 2 Quality assessment of the included literature

Item Item Item Item Item Item Item Item Item Item Item
Author/year 1 2 3 4 5 6 7 8 9 10 11 Score Quality

Wang/2017 √ √ × × • √ √ √ √ √ • 7 Medium
Jing/2018 √ √ √ × • √ √ √ √ × • 7 Medium
Zurawiecka/2018 √ √ √ × • √ √ √ √ × • 7 Medium
Rafique/2018 √ √ √ × • √ • √ √ √ • 7 Medium
Fernández-­Martínez √ √ √ × • √ √ √ √ × • 7 Medium
/2018
Wang/2019 √ √ √ × • √ • √ √ √ • 7 Medium
Zheng/2020 √ √ × × • √ • √ √ √ • 6 Medium
Jiang/2020 √ √ × × • √ • √ • √ • 5 Medium
Hu/2020 √ √ √ × • √ √ √ √ × • 7 Medium
Hashim/2020 √ √ √ × • √ √ √ √ √ • 8 High
Shellasih/2020 √ √ √ × • √ • √ × √ • 6 Medium
Karout/2021 √ √ √ × • √ • √ √ √ • 7 Medium

Note: √: yes; ×: no; •: unclear; Item1: Define the source of information (survey, record review); Item 2: List inclusion and exclusion criteria for exposed
and unexposed participants (cases and controls) or refer to previous publications; Item 3: Indicate time period used for identifying patients; Item 4:
Indicate whether or not participants were consecutive if not population-­based; Item 5: Indicate if evaluators of subjective components of study were
masked to other aspects of the status of the participants; Item 6: Describe any assessments undertaken for quality assurance purposes (eg, test/
retest of primary outcome measurements); Item 7: Explain any patient exclusions from analysis; Item 8: Describe how confounding was assessed
and/or controlled; Item 9: If applicable, explain how missing data were handled in the analysis; Item 10: Summarize patient response rates and
completeness of data collection; Item 11: Clarify what follow-­up, if any, was expected and the percentage of patients for which incomplete data or
follow-­up was obtained; Medium: 4–­7 (score); High: 8–­11 (score).

TA B L E 3 Results of meta-­analysis of the relationship between BMI and primary dysmenorrhea

Effect size Heterogeneity test Publication bias

Number of
Category studies OR (95% CI) p I2 (%) p Egger's test Begg's test
7,23–­33
Underweight 12 1.43 (1.18–­1.73) <0.001 78.9 <0.001 0.771 0.732
Overweight 127,23–­33 1.04 (0.85–­1.27) 0.709 65.9 0.001 0.230 0.837
7,26,28,33
Obesity 4 1.41 (0.76–­2.64) 0.280 27.7 0.246 0.346 0.308

Abbreviations: BMI, body mass index; CI, confidence interval OR, odds ratio.

in six countries (China, Poland, Arabia, Spain, Indonesia, and to 24.9 kg/m2, overweight from 25 kg/m2 to 29.9 kg/m2, and obesity
Lebanon), and most of the studies23,24,28–­31 were conducted was BMI of 30 kg/m2 or more. The quality assessment results of the
in China. Corresponding countries to their continents, 10 included studies are shown in Table 2. The quality assessment scores
studies7,23,24,26,28–­33 were conducted in Asia, and only two stud- of the included studies ranged from 5 to 8; 11 of the studies23–­33 were
25,27
ies were conducted in Europe. of medium quality, and one study7 was of high quality. The results of
Ten included studies23–­31,33explicitly mentioned the assessment of Begg's test and Egger's test revealed that there was no significant pub-
pain in primary dysmenorrhea, but two studies7,32 did not specifically lication bias (p > 0.05) in the association between each abnormal cate-
report it. A visual analog scale was used in four studies27,29,31,33 to as- gory of BMI (underweight, overweight, obesity) and the occurrence of
sess pain, which was simple and easy to use. In addition, two studies23,24 primary dysmenorrhea in the included studies (Table 3).
28,30
assessed pain through the appropriate guidelines, two studies
assessed pain through content in the textbook, and two studies25,26
used other scales separately. All included studies covered the following 3.2 | Synthesis: relationships
three categories of BMI: underweight, normal weight, and overweight;
only four studies7,26,28,33 explicitly included the category of obesity. Table 3 showed the results of the relation between each cat-
Although the cut-­off values of each category of BMI varied among egory of BMI and the occurrence of primary dysmenorrhea. Twelve
the included studies, in most studies, the criteria for being under- studies7,23–­33 reported the relation between being underweight and
2 2
weight was BMI less than 18.5 kg/m , normal weight from 18.5 kg/m the occurrence of primary dysmenorrhea, involving a total of 6545
WU et al. | 1369

F I G U R E 2 Forest plot of the relation between underweight and primary dysmenorrhea

TA B L E 4 Results of subgroup analysis

Effect size Heterogeneity test
of the relation between underweight, and Number of
overweight, and primary dysmenorrhea Subgroups studies OR (95% CI) p I2 (%) p

Underweight <0.001
7,23,24,26,28–­33
Asia 10 1.35 (1.12–­1.64) 0.002 79.9 <0.001
Europe 225,27 2.37 (1.09–­5.17) 0.030 49.2 0.161
Overweight 0.709
Asia 107,23,24,26,28–­33 0.89 (0.80–­1.19) 0.820 63.0 0.004
25,27
Europe 2 1.56 (0.67–­3.64) 0.300 62.6 0.102

Abbreviations: CI, confidence interval; OR, odds ratio.

underweight women (Figure 2). We used a random-­effects model to There was a regional difference in cut-­off value of the overweight
calculate the OR along with its 95% CI due to the heterogeneity test category of BMI in the included literature in this study. Eight included
testified p < 0.001 and I2 = 78.9%. The results (Table 3) suggested that studies23–­25,27,29–­32 covered only three categories (underweight, nor-
being underweight may be associated with the occurrence of primary mal weight, overweight) of BMI and did not further distinguish be-
dysmenorrhea (OR 1.43; 95% CI 1.18–­1.73). In addition, the results of tween overweight and obesity. In order to explore the association
the subgroup analysis (Table 4) indicated that the underweight women between overweight and the occurrence of primary dysmenorrhea,
in Europe25,27 (OR 2.37; 95% CI 1.09–­5.17) might be more likely to ex- we ignored the regional difference mentioned above and roughly
perience primary dysmenorrhea than Asian women7,23,24,26,28–­33 (OR considered the category of BMI in the above-­undifferentiated condi-
1.35; 95% CI 1.12–­1.64). tion to be overweight. As a result, a total of 12 studies7,23–­33 were
1370 | WU et al.

involved and the corresponding participants included in these stud- dysmenorrhea. However, overweight and obesity might not be as-
ies numbered 3098. (Figure 3). Since the heterogeneity test reported sociated with the occurrence of primary dysmenorrhea. In addition,
p = 0.001 and I2 = 65.9%, a random-­effects model was used. The find- the subgroup analysis showed that underweight European women
ings (Table 3) suggested that being overweight may not be associated may be more likely to experience primary dysmenorrhea than un-
with the occurrence of primary dysmenorrhea. In subgroup analyses derweight Asian women. But there appeared to be no difference
(Table 4), we found that overweight European women may not differ between overweight European women compared with overweight
from overweight Asian women in their association with the occur- Asian women in their relation with the occurrence of primary
rence of primary dysmenorrhea compared with overweight women. dysmenorrhea.
Only four studies7,26,28,33 explicitly explored the relation be- There are several previous10,11,34 studies that support the find-
tween obesity and the occurrence of primary dysmenorrhea, cor- ings of this study regarding the association between underweight
responding to a sample size of 94 (Figure 4). The heterogeneity test and primary dysmenorrhea. Although the cause of primary dys-
showed p = 0.246 and I2 = 27.7%, so we used a fixed-­effects model menorrhea is still not fully understood, the currently accepted
and did not perform the relevant subgroup analyses. The results of pathogenesis for the occurrence of primary dysmenorrhea is the
the meta-­analysis (Table 3) indicated that obesity might not be re- overproduction of prostaglandins. Once too much prostaglandin
lated to the development of primary dysmenorrhea. is released, the uterus will contract excessively, causing increased
pressure and reduced blood flow in the uterus, leading to ischemia
and hypoxia, which result in the occurrence of primary dysmenor-
4 | DISCUSSION rhea.4 Being underweight means that women may have low body
fat and suffer from malnutrition. A certain amount of body fat is im-
In this study, we identified the relation between BMI and primary portant for the maintenance of a normal ovulation cycle,35 but low
dysmenorrhea by analyzing the relevant included studies. We found body fat may interfere with normal ovulation and menstrual cycles,
that being underweight may increase the risk of developing primary which might lead to excessive prostaglandin release,11 and so cause

F I G U R E 3 Forest plot of the relation between overweight and primary dysmenorrhea

WU et al. | 1371

Study, year (n) OR (95% CI) Weight

Rafique, 2018 (41) 5.33 (1.22, 23.42) 17.96

Wang, 2019 (10) 0.84 (0.22, 3.29) 21.18

Hashim, 2020 (24) 0.97 (0.34, 2.75) 35.98

Karout, 2021 (19) 1.44 (0.41, 5.08) 24.88

2
Overall, IV (I = 27.7%, p = 0.246) 1.41 (0.75, 2.64) 100.00

.03125 1 32

F I G U R E 4 Forest plot of the relation between obesity and primary dysmenorrhea

excessive uterine contractions, leading to the occurrence of primary regional differences, and cultural differences. In addition, the cut-­off
dysmenorrhea. In addition, underweight women produced less es- values for the overweight category of BMI were not consistent among
trogen and this may also lead to irregularity in their ovulatory cycles the included studies exploring the relation between overweight and
and increase the risk of the occurrence of primary dysmenorrhea.9,36 the occurrence of primary dysmenorrhea. And some included studies
Related research showed that malnutrition was a possible factor for did not clearly distinguish between overweight and obesity. So it is
the occurrence of primary dysmenorrhea.37 The above explanation possible that the overweight category of BMI may have also included
may support the finding that being underweight may increase the obese women. It can be seen that the difference in the cut-­off values
risk of the occurrence of primary dysmenorrhea. It has been sug- of overweight may have had a significant effect on the results of this
gested that Europeans were likely to have a lower amount of body fat study. Therefore, the World Health Organization classification criteria
than Asians,38 which might explain why the underweight European for BMI should be used whenever possible in future relevant studies,
women were more likely than Asian women to experience primary and a distinction should be made between the overweight category
dysmenorrhea. Moreover, sociocultural factors and environment in and the obese category.
different continents may also be responsible for this result. More Tembhurne and Mitra42 and Abadi Bavil et al43 pointed out that
studies are needed to provide a more scientific interpretation of the obesity may increase the risk of developing primary dysmenorrhea.
subgroup results mentioned above. Temur et al44 further believed that obesity may be one of the cor-
39 40
Al-­Matouq et al and Nloh et al found that there appeared to be rectable influencing factors for primary dysmenorrhea. The results
no relation between overweight and the development of primary dys- of the above studies are not consistent with this study. In our study,
menorrhea, which is consistent with the results of this meta-­analysis. only four studies explored the relation between obesity and the oc-
26 41
However, Rafique and Al-­Sheikh and Elizondo-­Montemayor et al currence of primary dysmenorrhea, and the sample sizes involved in
found that there was a significant relation between overweight and these studies were small. Moreover, the cut-­off values for the obe-
primary dysmenorrhea in their studies. The contrary results of the sity of BMI in the included studies were not identical. These factors
studies may be related to factors such as the greatly varied sample size, may have contributed to the inconsistency between the results of
1372 | WU et al.

this study and those of other available studies. As far as we know, between each BMI category and the occurrence of primary dysmen-
obesity may increase the production of both prostaglandin and orrhea was not fully clarified in this study, maintaining a balanced
estrogen, which would lead to excessive uterine contraction and diet and an appropriate lifestyle is beneficial for people to have the
abnormal ovulatory cycles.42,45 So obesity may be associated with normal category of BMI and live a healthy life, which may play a role
the development of primary dysmenorrhea and it was worth noting in preventing the occurrence of primary dysmenorrhea.
that the prevalence of obesity among adolescents is increasing glob-
ally.46 If obesity does increase the risk of primary dysmenorrhea to AU T H O R C O N T R I B U T I O N S
some extent, then the incidence of primary dysmenorrhea in young Concept and design of the article: Lingsha Wu, Haiyan Fang; Search
women will also increase year by year. Based on these findings, it strategy and selection criteria: Lingsha Wu, Jie Tang, Haiyan Fang;
is necessary to conduct more large-­sample studies in the future to Data extraction: Lingsha Wu, Jing Zhang; Quality assessment: Lingsha
clarify the relation between obesity and the occurrence of primary Wu, Jing Zhang, Jie Tang, Haiyan Fang; Statistical analyses: Lingsha
dysmenorrhea and we should use cut-­off values for the obesity cat- Wu, Jing Zhang; Manuscript writing: Lingsha Wu; All authors ap-
egory of the World Health Organization classification criteria. proved the final version of the manuscript.
Several limitations of the current meta-­analysis should be de-
scribed. First, the included studies were all cross-­sectional studies F U N D I N G I N FO R M AT I O N
and the sample sizes varied considerably between studies, which may This systematic review was funded by the Key Project of Natural
produce some bias. We should therefore interpret the relevant results Science Research Fund for Universities of Anhui Province Education
with caution. Second, we ignored the regional difference in the cut-­off Department (KJ2020A0444).
value of the overweight category of BMI in the included studies and
roughly classified the category of BMI that did not further distinguish C O N FL I C T O F I N T E R E S T
between overweight and obesity as overweight. This may lead to some None.
deviation in the results. Third, because of the small amount of litera-
ture included in this study, only subgroup analyses of continents were ORCID
ultimately performed, which could also affect the results of the rele- Lingsha Wu https://orcid.org/0000-0002-5137-7057
vant meta-­analysis. Fourth, potential confounding factors, such as the
family history of primary dysmenorrhea, length of menstrual cycle, and REFERENCES
dietary habits, could not be completely excluded. These would prob- 1. Coco AS. Primary dysmenorrhea. Am Fam Physician.
ably affect the results. In addition to the above limitations, this study 1999;60:489-­496.
2. Kho KA, Shields JK. Diagnosis and management of primary dys-
has the following strengths. First, we searched 11 medical databases,
menorrhea. Jama. 2020;323:268-­269.
supplemented by manual searching to make sure the studies were re- 3. Giletew A, Bekele W. Prevalence and associated factors of primary
trieved completely. Second, we carried out quality control, which was dysmenorrhea among Debre Tabor University students, north
performed by two or three researchers independently in terms of study Central Ethiopia. Int J Biomed Eng Clin Sci. 2019;4:70-­74.
4. Ferries-­Rowe E, Corey E, Archer JS. Primary dysmenorrhea: diag-
selection, data extraction, and study quality assessment. Furthermore,
nosis and therapy. Obstet Gynecol. 2020;136:1047-­1058.
we found no publication bias in the included studies by Begg's test and 5. Chen Y, Tian S, Tian J, Shu S. Wrist-­ankle acupuncture (WAA) for pri-
Egger's test and we used subgroup analyses to explore and control the mary dysmenorrhea (PD) of young females: study protocol for a ran-
sources of heterogeneity. In addition, we separated the categories of domized controlled trial. BMC Complement Altern Med. 2017;17:421.
6. Tomás-­Rodríguez MI, Palazón-­Bru A, Martínez-­St John DR,
BMI, thus creating an opportunity to explore each category of BMI for
Navarro-­Cremades F, Toledo-­Marhuenda JV, Gil-­Guillén VF.
its association with the occurrence of primary dysmenorrhea. Factors associated with increased pain in primary dysmenorrhea:
analysis using a multivariate ordered logistic regression model.
J Pediatr Adolesc Gynecol. 2017;30:199-­202.
5 | CO N C LU S I O N 7. Hashim RT, Alkhalifah SS, Alsalman AA, et al. Prevalence of primary
dysmenorrhea and its effect on the quality of life amongst female
medical students at King Saud University, Riyadh, Saudi Arabia. A
The meta-­analysis suggested that being underweight may increase Cross-­Sectional Study Saudi Med J. 2020;41:283-­289.
the risk of the occurrence of primary dysmenorrhea, whereas being 8. Yu K, Liu X, Alhamzawi R, Becker F, Lord J. Statistical methods
overweight and obese may not be associated with the development for body mass index: a selective review. Stat Methods Med Res.
2018;27:798-­811.
of primary dysmenorrhea. Furthermore, the underweight women in
9. Banu SH, Razick JA. Study on the menstruation pattern and weight
Europe may be more likely than Asian women to suffer from pri- status of college girls in Chennai, South India. Online J Health Allied
mary dysmenorrhea. The cut-­off values for each category of BMI Sci. 2020;19:4.
included in this study were not completely consistent, which may 10. Mohapatra D, Mishra T, Behera M, Panda P. A study of relation be-
tween body mass index and dysmenorrhea and its impact on daily ac-
have influenced the results of this study to some extent. This should
tivities of medical students. Asian J Pharm Clin Res. 2016;9:297-­299.
be avoided in future studies and more high-­quality studies with 11. Çinar GN, Akbayrak T, Gürşen C, et al. Factors related to primary
large samples should be conducted to explore the relation between dysmenorrhea in Turkish women: a multiple multinomial logistic re-
BMI categories and primary dysmenorrhea. Although the relation gression analysis. Reprod Sci. 2021;28:381-­392.
WU et al. | 1373

12. Shahid A, Tatiq M, Sulaman H, Shaista BS, Habib SA. Frequency and university students: a cross-­sectional study. J Pediatr Adolesc
severity of primary dysmenorrhea in adolescent females. Pak. J Med Gynecol. 2020;33:15-­22.
Sci. 2020;14:1952-­1954. 32. Shellasih NM, Ariyanti F. Factors of primary dysmenorrhea in junior
13. Shahoei R, Nouri B, Darvishi N, et al. Prevalence of menstrual high school students in South Tangerang City, Indonesia, 2018.
disorders and its related factors in the students of Kurdistan J Public Hlth Dev. 2020;18:73-­78.
University of Medical Science in 2018. Sci J Kurdistan Univ Med Sci. 33. Karout S, Soubra L, Rahme D, Karout L, Khojah HMJ, Itani R.
2020;25:31-­41. Prevalence, risk factors, and management practices of primary
14. Jiang W, Hua XG, Hu CY, Li FL, Huang K, Zhang XJ. The prev- dysmenorrhea among young females. BMC Womens Health.
alence and risk factors of menstrual pain of married women 2021;21:1-­14.
in Anhui Province, China. Eur J Obstet Gynecol Reprod Biol. 34. Khalid M, Jamali T, Ghani U, Shahid T, Ahmed T, Nasir T. Severity
2018;229:190-­194. and relation of primary dysmenorrhea and body mass index in un-
15. Vilšinskaitė DS, Vaidokaitė G, Mačys Ž, Bumbulienė Ž. The dergraduate students of Karachi: a cross sectional survey. J Pak
risk factors of dysmenorrhea in young women. Wiad Lek. Med Assoc. 2020;70:1299-­1304.
2019;72:1170-­1174. 35. Ju H, Jones M, Mishra GD. A U-­shaped relationship between body
16. Barcikowska Z, Wójcik-­Bilkiewicz K, Sobierajska-­Rek A, mass index and dysmenorrhea: a longitudinal study. PLoS One.
Grzybowska ME, Wąż P, Zorena K. Dysmenorrhea and associated 2015;10:e0134187.
factors among polish women: a cross-­sectional study. Pain Res 36. Singh M, Rajoura OP, Honnakamble RA. Menstrual patterns and
Manag. 2020;2020:6161536. problems in association with body mass index among adolescent
17. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for school girls. J Family Med Prim Care. 2019;8:2855-­2858.
reporting systematic reviews and meta-­analyses of studies that 37. Rad M, Sabzevari MT, Rastaghi S, Dehnavi ZM. The relationship be-
evaluate healthcare interventions: explanation and elaboration. tween anthropometric index and primary dysmenorehea in female
BMJ. 2009;339:b2700. high school students. J Educ Health Promot. 2018;7:34.
18. Zeng X, Zhang Y, Kwong JS, et al. The methodological quality as- 38. Di Angelantonio E, Bhupathiraju SN, Wormser D, et al. Body-­mass
sessment tools for preclinical and clinical studies, systematic re- index and all-­c ause mortality: individual-­participant-­data meta-­
view and meta-­analysis, and clinical practice guideline: a systematic analysis of 239 prospective studies in four continents. Lancet.
review. J Evid Based Med. 2015;8:2-­10. 2016;388:776-­786.
19. Wang J, You D, Wang H, et al. Association between homocysteine 39. Al-­Matouq S, Al-­Mutairi H, Al-­Mutairi O, et al. Dysmenorrhea
and obesity: a meta-­analysis. J Evid Based Med. 2021;14:208-­217. among high-­school students and its associated factors in Kuwait.
20. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-­ BMC Pediatr. 2019;19:1-­12.
analysis. Stat Med. 2002;21:1539-­1558. 40. Nloh AM, Ngadjui E, Vogue N, et al. Prevalence and factors associated
21. Higgins JP, Thomas J, Chandler J, et al. Cochrane Handbook for with dysmenorrhea in women at child bearing age in the Dschang
Systematic Reviews of Interventions. John Wiley & Sons; 2019. Health District, West-­Cameroon. Pan Afr Med J. 2020;37:178.
Accessed April 26, 2022. https://train​ing.cochr​ane.org/handbook 41. Elizondo-­Montemayor L, Hernández-­Escobar C, Lara-­Torre E,
22. Egger M, Davey Smith G, Schneider M, Minder C. Bias in Nieblas B, Gómez-­C armona M. Gynecologic and obstetric conse-
meta-­analysis detected by a simple, graphical test. BMJ. quences of obesity in adolescent girls. J Pediatr Adolesc Gynecol.
1997;315:629-­634. 2017;30:156-­168.
23. Wang XP. Analysis on TCM Syndromes and Related Factors of Primary 42. Tembhurne S, Mitra M. Relationship between body mass compo-
Dysmenorrhea Among Female College Students in Changsha. Master's sition and primary dysmenorrhoea. Ind J Physioth Occupat Therapy.
thesis. Hunan University of Chinese Medicine; 2017. 2016;10:76-­81.
24. Jing GL. Study on the Current Situation and Influencing Factors of 43. Abadi Bavil D, Dolatian M, Mahmoodi Z, Akbarzadeh BA. A com-
Dysmenorrhea of Female Students in an University. Master's the- parison of physical activity and nutrition in young women with and
sis. Zhengzhou University; 2018. without primary dysmenorrhea. F1000Res. 2018;7:59.
25. Zurawiecka M, Wronka I. Association of primary dysmenorrhea 44. Temur M, Balci UG, Güçlü YA, et al. The relationship between obe-
with anthropometrical and socio-­economic factors in polish uni- sity and primary dysmenorrhea: does increase in body mass index
versity students. J Obstet Gynaecol Res. 2018;44:1259-­1267. effect dysmenorrhea? J Turk Ger Gynecol Assoc. 2016;17:S217.
26. Rafique N, Al-­Sheikh MH. Prevalence of primary dysmenorrhea 45. Tang Y, Zhao M, Lin L, et al. Is body mass index associated with the
and its relationship with body mass index. J Obstet Gynaecol Res. incidence of endometriosis and the severity of dysmenorrhoea: a
2018;44:1773-­1778. case-­control study in China? BMJ Open. 2020;10:e037095.
27. Fernández-­Martínez E, Onieva-­Zafra MD, Parra-­Fernández ML. 46. Abarca-­Gómez L, Abdeen ZA, Hamid ZA, et al. Worldwide trends in
Lifestyle and prevalence of dysmenorrhea among Spanish female body-­mass index, underweight, overweight, and obesity from 1975
university students. PLoS One. 2018;13:e0201894. to 2016: a pooled analysis of 2416 population-­based measurement
28. Wang HR, Zhang K, Gao X, Luo YY, Zhang L. Survey of primary studies in 128·9 million children, adolescents, and adults. Lancet.
dysmenorrhea incidence and its influencing factors among fe- 2017;390:2627-­2642.
male students of a university in Ningxia. J Ningxia Med Univ.
2019;41:1055-­1059.
29. Zheng W, Hao X, Li W, Liu JX. Epidemiological investigation and
correlation analysis of dysmenorrhea among middle school stu- How to cite this article: Wu L, Zhang J, Tang J, Fang H.
dents in Qingdao. Shandong J Tradit Chin Med. 2020;39:688-­692. The relation between body mass index and primary
30. Jiang W. The Prevalence and Correlated Factors of Primary dysmenorrhea: A systematic review and meta-analysis. Acta
Dysmenorrhea Among Rural Women of Childbearing Age. Master's
Obstet Gynecol Scand. 2022;101:1364-1373. doi: 10.1111/
thesis. Anhui Medical University; 2020.
31. Hu Z, Tang L, Chen L, Kaminga AC, Xu H. Prevalence and risk fac- aogs.14449
tors associated with primary dysmenorrhea among Chinese female