Psychosocial Factors at Work and Menstrual Disorders: a Protocol for a Systematic Review and Meta-analysis

Introduction: Menstrual abnormalities, such as menstrual cycle disorders or related symptoms (e.g., pelvic pain, premenstrual syndrome [PMS], menopausal symptoms), affect female workers’ quality of life and work capacity. The fertility of female workers is also a worldwide concern. However, the association between psychosocial factors at work and menstrual abnormalities or fertility is not well understood. To address this relationship, we will conduct a systematic review and a meta-analysis of the literature that has utilized a longitudinal or prospective cohort design. Methods and analysis: The inclusion criteria for this systematic review and meta-analysis are dened as follows: (P) Adult female workers (over 18 years old), (E) Presence of adverse psychosocial factors at work, (C) Absence of adverse psychosocial factors at work, and (O) Any menstrual cycle disorders, menstrual-related symptoms, or fertility. MEDLINE, EMBASE, PsycINFO, PsycARTICLES, and Japan Medical Abstracts Society electronic databases will be used to search for published studies. The statistical synthesis of the studies included in the meta-analysis will be conducted to estimate pooled coecients and 95% CIs. We will use the Risk of Bias in Non-randomized Studies-of Interventions (ROBINS-I) to determine the quality of selected studies. To assess meta-bias, Egger’s test, along with a funnel plot, will be used to check for publication bias. Lastly, we will examine heterogeneity using the χ² test with Cochran’s Q statistic and I 2 . Ethics and dissemination: The results and ndings will be submitted and published in a scientic peer-reviewed journal and will be disseminated broadly to researchers and policymakers interested in the translatability of scientic evidence into good practices. Trial registration: The study protocol was registered at the UMIN registry (registration number: UMIN000039488). The registration date is 14th Feb 2020. URL:


Introduction
The female working participation rate has been increasing along with improved educational opportunities and the lack of labor force [1]. Poor reproductive health of working women, including infertility and menstrual abnormalities, has a major effect on health and work outcomes throughout the preconception period (before pregnancy) to the climacteric period [2]. Reproductive health among female workers can be a social issue as well as affect females' career choices.
Menstrual abnormalities can be differentiated into menstrual cycle disorders and associated menstrual symptoms. Menstrual cycle disorders, including shorter or longer cycles in reproductive age, are related to infertility [3][4][5], breast and ovarian cancer [6,7], diabetes [8], and cardiovascular disease [9,10]. In addition to age, weight, tumor, in ammation, or endocrine dysfunctions, psychosocial factors can affect the menstrual cycle. In general, the menstrual cycle is regulated by hypothalamic gonadotropin-releasing hormone (GnRH); however, the function and secretion of GnRH can be inhibited by hormones released by the hypothalamic-pituitary-adrenal (HPA) axis, which is activated in response to stressors [11,12]. The stress reactions can directly change the level of serum concentrations of sex hormones in response to psychosocial factors [13]. It may affect fertility outcomes, such as contraception or pregnancy. Female workers facing various types of stressors may be at risk of menstrual cycle disorders and infertility.
Alongside the menstrual cycle disorders and adverse fertility outcomes, menstrual-related symptoms (e.g., pelvic pain, premenstrual syndrome [PMS], menopausal symptoms) have a considerable effect on workers' quality of life (QOL) and diminished work capacity [14][15][16]. The cost of productivity loss was reported $15,737 per working woman per year due to absenteeism and presenteeism for endometriosis alone, which is a major cause of heavy pelvic pain [17]. A longitudinally designed study revealed that PMS among female workers was associated with higher absenteeism and less work productivity [18]. Furthermore, the high prevalence of menstrual-related symptoms has been reported worldwide, in particular, chronic pelvic pain (CPP); 24% [19], PMS; 30-40%, premenstrual dysphoric disorder (PMDD) (a more serious variant of PMS) 3-8% [20].
Cytokines/chemokines and muscle contractions are a direct cause of menstrual pain [21]. Still, psychosocial factors at work, such as shift work, low support, or high demand, may play a role in alleviating or aggravating this pain. Stress reactions can also strengthen PMS through di culty in mood regulation and increasing sensitivity to stressful experiences [22,23]. Regarding vasomotor symptoms in perimenopause (e.g., hot ushes), psychosocial factors can be in charge of exacerbating symptoms [24]. Therefore, it is important to identify psychosocial factors (i.e., stressors) at work that affect menstrualrelated symptoms and explore the mechanisms through which they exert this effect.
Occupational psychosocial environments thus have a non-neglectable effect on female menstrual health. Previous literature reviews have revealed some physical and chemical workplace factors (e.g., chemical exposure, low temperature at work, lifting heavyweight) related to menstrual cycle disorders, dysmenorrhea, and reproductive outcome [25,26]. However, limited evidence has supported the relationship between psychosocial factors at work and menstrual outcomes. Accumulating evidence indicates the adverse health effects of shift work. The meta-analysis of Stocker et al. showed that shift work increased the risk of menstrual disruption and infertility [27]. Shiftwork can contribute to cycle disorder by changing the gonadotropin secretion cycle as a result of circadian rhythm disturbances as well as stress-related dysfunction of the hypothalamus-pituitary-ovarian axis [28,29]. Concerning menstrual-related symptoms, a cross-sectional survey in the workplace revealed that low job control, low co-worker support, and low job security were found to be associated with a higher risk for menstrual pain [30]. Somatic symptoms in perimenopause are also known to be affected by work-related psychosocial factors, such as supervisor support [31].
However, psychosocial factors in the workplace associated with menstrual abnormalities in female workers have not been comprehensively investigated. Moreover, no meta-analysis with well-designed longitudinal studies has been conducted to our knowledge.

Objectives
This systematic review and a meta-analysis study aims to investigate the comprehensive association between psychosocial factors at work and menstrual-related disorders/symptoms using longitudinal data.

Study design
This systematic review and meta-analysis of longitudinal studies will be conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) guidelines [32] and reported following the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines [33]. The study protocol was registered at the UMIN registry (registration number: UMIN000039488). The registration date is 14th Feb 2018. The PRISMA-P checklist can be available in Supplemental appendix 1.

PECO and Eligibility criteria
For this systematic review and meta-analysis, the eligible participants, exposures, comparisons, and outcomes (PECO) of the studies will include: (P) Adult female workers (over 18 years old).
(E) Presence of adverse psychosocial factors at work.
(C) Absence of adverse psychosocial factors at work.
(O) Any menstrual cycle disorders, menstrual-related symptoms, or fertility.
Job strain, effort-reward imbalance, working hours, shift work, low social support, and other organizational-level factors, work conditions, and interpersonal relationships at work will be included as adverse psychosocial factors.
Inclusion criteria are as follows: (1) studies that included female participants who were working as of baseline survey period.
(2) studies which assessed the adverse psychosocial factors at work as exposure variables at baseline survey.
(3) studies that assessed any menstrual-related outcome at baseline and follow-up surveys. Exclusion criteria are as follows: (1) studies targeting pregnancy-related outcomes (e.g., premature birth).
Information sources and search strategy MEDLINE, EMBASE, PsycINFO, PsycARTICLES, and Japan Medical Abstracts Society electronic databases will be used to search for published studies. The search terms will include words related to the PECO of the studies (Supplementary Appendix 2). The search terms for psychosocial factors at work will be de ned based on the previous meta-analysis [34][35][36]. The search terms for outcome variables will be carefully selected by reviewing current evidence regarding psychological factors at work and menstrualrelated health. The authors consulted a public health medical doctor specializing in obstetrics and gynecology (outside the research team) to con rm the search terms. Infectious or in ammatory diseases and pelvic organ prolapse were not included in outcomes. Terms of menstrual-related outcomes were retrieved from previous systematic reviews [17,27,[37][38][39].

Study records
Microsoft Excel (Washington, USA) will be used to manage the data. Before screening the studies, NS will remove duplicate entries from the Excel le.

Selection process
Nineteen investigators (NS, KI, KW, EM, HE, AI, KT, YH, YKom, MIi, YO, ASa, YA, MIw, YKob, RI, ASh, AT, and NK) will screed the studies independently according to the eligibility criteria. After excluding duplicate records, the remaining articles will be distributed to 16 investigators (NS, KI, KW, EM, HE, AI, KT, YH, YKom, MIi, YO, ASa, YA, MIw, YKob, and RI). Two investigators will receive the same set of articles. Each investigator will screen the title and abstract of each article independently to select eligible studies according to the eligibility criteria (sifting phase). In this phase, we will gather the full texts of all eligible studies. Subsequently, two investigators will review the full texts independently. Any disagreements will be settled by consensus among all authors, and the reasons for excluding studies will be documented. A ow chart of the entire review process will be created.

Data collection process
Data will be extracted from the included studies independently by 16 investigators (NS, KI, KW, EM, HE, AI, KT, YH, YKom, MIi, YO, ASa, YA, MIw, YKob, and RI) using a standardized data extraction form. The authors will discuss any disagreements or inconsistencies until a consensus is achieved. From the included articles, the investigators will extract information on publication year, study design, the country of study origin, the number of participants completing the baseline survey and included in the statistical analysis, demographic characteristics of the participants (i.e., age, occupation), the length of follow-up and attrition rate, exposure variables (i.e., adverse psychosocial factors at work), outcome variables (i.e., menstrual abnormalities, fertility), and data necessary to calculate the coe cients (β, γ), odds ratios (ORs), relative risks (RRs) or hazard ratios (HRs) with standard errors (SEs) or 95% con dence intervals (CIs) to determine the association between psychosocial factors at work and menstrual abnormalities or fertility. In case of missing data, we will try to contact the authors of the included studies to obtain missing information.

Data synthesis
The included studies will be integrated in a meta-analysis and strati ed by types of measures of association (β, γ, OR, RR, and HR). Subsequently, pooled coe cients and 95% CIs will be estimated. For the studies that reported ORs, RRs, or HRs, we will calculate log-transformed ORs, RRs, or HRs and determine SEs based on 95% CIs. A funnel plot and Egger's test will be used to plot these parameters will to examine publication bias.

Risk of bias in individual studies and assessment of metabias
The 16 investigators (NS, KI, KW, EM, HE, AI, KT, YH, YKom, MIi, YO, ASa, YA, MIw, YKob, and RI) will utilize the internationally recognized tool for evaluating the risk of bias (Risk of Bias in Non-Randomized Studies -of Interventions; ROBINS-I) to independently assess the quality of included studies [40]. The risk of bias will be classi ed as low, high, or unclear. Any discrepancies between the investigators in the quality assessment will be documented and discussed until achieving a consensus.

Primary analyses
For the main analysis, we will synthesize all types of psychosocial factors at work and all types of menstrual-related disorders/symptoms. The outcomes assessed on a continuous scale will be converted to dichotomous variables based on reasonable and theoretically sound cut-off points. Dichotomous and continuous variables for which no reasonable cut-off point could be determined will be analyzed separately.
At least three eligible studies will have to be gathered to conduct a meta-analysis; otherwise (i.e., if only one or two studies will be eligible and included), the results will be presented in a narrative format. A xed-effect model will be used with homogeneous data; otherwise, a random-effects model will be used.
The chi-square test with Cochran's Q statistic and I 2 will be used to test heterogeneity. I 2 Values of 25%, 50%, and 75% indicate low, medium, and high heterogeneity, respectively.

Subgroup and sensitivity analyses
Subgroup analyses will be conducted to compare the results under a speci c type of exposure (e.g., night shift, high job strain) and outcome variables (e.g., menstrual cycle disorder, PMS), if we have enough data to conduct such analyses. Meta-regression will be conducted in case of signi cant pooled associations with any grouping characteristics. A sensitivity analysis will be run for studies in which the ROBINS-I is classi ed as low risk [40].

Patient and public involvement
There is no direct patient or public involvement in the design of this study.

Ethics And Dissemination
Since our data will be extracted from published studies, and thus privacy issues will not be of concern, ethical approval will not be needed to apply this review protocol. Results and The ndings will be submitted to a scienti c peer-reviewed journal for publication and disseminated broadly to researchers and policymakers interested in the translatability of scienti c evidence into good practices.

Strengths and limitations
This will be the rst systematic review and meta-analysis to explore the comprehensive effects of a wide range of psychosocial factors at work on menstrual-related disorders among female workers. The ndings of this study will establish the link between psychosocial factors at work and menstrual abnormalities or fertility by restricting the eligible studies to those that utilized longitudinal or prospective cohort design. Thus, reviewing the current evidence would contribute to promoting women's health at the