Characteristics of the gut microbiota in women with premenstrual symptoms: A cross-sectional study.

PURPOSE: Premenstrual symptoms can negatively impact the quality of life of women through a range of mood, behavioral, and physical symptoms. The association between the microbiota and brain function has been extensively studied. Here, we examined the characteristics of the microbiota in women with premenstrual disorders (PMDs) and the association between premenstrual symptoms and the microbiota.
MATERIALS AND METHODS: In this single center cross-sectional pilot study, we recruited 27 women reporting premenstrual symptoms and 29 women with no serious premenstrual symptoms. Among them, we further selected 21 women experiencing premenstrual symptoms resulting in interference to their social life (PMDs group) and 22 women with no serious premenstrual symptoms and thereby no interference to their social life (control group). The severity of symptoms was evaluated by a premenstrual symptoms questionnaire (PSQ). Inflammatory markers were analyzed in blood samples, including C reactive protein, soluble CD14, and lipopolysaccharide binding protein. Sequencing of 16S ribosomal ribonucleic acid genes was performed on stool samples.
RESULTS: Inflammatory markers in blood samples did not differ significantly between the PMDs and control groups. A difference in beta, but not alpha diversity, was detected for the gut microbiotas of the PMDs and control groups. The relative abundance of the Bacteroidetes phylum was lower in the PMDs group. At the genus level, the prevalence was decreased for Butyricicoccus, Extibacter, Megasphaera, and Parabacteroides and increased for Anaerotaenia in the PMDs group, but after false discovery rate correction, these differences were no longer significant. Linear discriminant effect size analysis revealed a decrease in Extibacter, Butyricicoccus, Megasphaera, and Parabacteroides and an increase in Anaerotaenia in the PMDs group. The PSQ total score correlated with Anaerotaenia, Extibacter, and Parabacteroides. Multiple regression analysis showed that Parabacteroides and Megasphaera negatively predicted the PSQ total score.
CONCLUSION: The properties of the gut microbiota are associated with premenstrual symptoms.

Introduction

Premenstrual symptoms encompass a range of psycho-physical symptoms observed before menstruation, which interfere with the quality of life of many women between menarche and menopause [1-3]. In epidemiologic surveys, the prevalence of premenstrual symptoms is high (80%–90%) [4]. As a disease, it has been classified as premenstrual syndrome (PMS) by the field of obstetrics and gynecology and as premenstrual dysphoric disorder (PMDD) by the field of psychiatry, but recently both have been recognized by the inclusive term, premenstrual disorders (PMDs) [5]. Various causes have been suggested, including hormonal changes, serotonergic dysfunction, impaired gamma-aminobutyric acid (GABA) function, stress, and poor lifestyle habits such as longer durations of internet use and shorter sleep durations, but the precise pathophysiology of PMDs remains unknown [6-9].

As “another organ”, the gut microbiota exhibits complex interactions with the immune, metabolic, and endocrine systems through the host’s intestinal epithelium, and maintains a delicate balance [10-12]. Many important diseases such as obesity, metabolic disease, cardiovascular disease, inflammatory diseases, and brain disorders are found to be associated with differences in the gut microbiome [13, 14]. Molecular methods, especially amplicon sequence analysis using next generation sequencing, have enabled us to detect the diversity and composition of the gut microbiota in detail.

Recently, the association between the microbiota and brain function, such as in the case of major depressive disorder (MDD), has been extensively studied [15-18]. The gut microbiota communicates with the brain through neuroendocrine, neuroimmune, and neural pathways, and this is known as the gut microbiota–brain axis [14, 19–21]. According to clinical data, MDD is associated with low-grade inflammation (C reactive protein (CRP) >3 mg/L) [22, 23]. One possible mechanism may be microbiota-related inflammation, or so-called “leaky gut” [24]. A dysfunctional intestinal barrier may permit the translocation of Gram-negative bacteria and the bacterial endotoxin (lipopolysaccharide, LPS) from the gut microbiota to the bloodstream [24]. Bacterial translocation induces LPS-binding protein (LBP) and soluble CD14 (sCD14) that potentiates inflammation through Toll-like receptor (TLR)-4 and NF-κB activation [24]. LBP and sCD14 may be markers of endotoxemia and are reported to be associated with depressive symptoms [25, 26].

PMDs are linked to various mood and behavioral symptoms, which overlap with MDD. For treatment, serotonin reuptake inhibitors are recommended and commonly prescribed for both PMDs and MDD [2, 27–29]. Despite the commonality between PMDs and MDD, there have been no reports to date profiling the microbiota in PMDs. Our goal in this pilot study was to examine the possibility that characteristics of the gut microbiota may be related to the pathogenesis of PMDs. The aims of this study were to: (1) test the association between PMDs and bacterial translocation, (2) study the characteristics of the gut microbiota in women with PMDs, and (3) study the association between the gut microbiota and premenstrual symptom severity.

Materials and methods

Ethics approval and informed consent

The study was carried out in accordance with the principles outlined in the Declaration of Helsinki. The trial protocol was approved by the Ethics Committee of Kindai University (approval number: 31–057). Written informed consent was obtained from all participants.

Settings and participants

This study was conducted at an obstetrics and gynecology outpatient clinic in Osaka, Japan. Study participants were enrolled between September 2019 and August 2020. We recruited patients who wished to receive treatment for their premenstrual symptoms (P group, n = 27) and healthy volunteers who did not report serious premenstrual symptoms (N group, n = 29) (Fig 1). The inclusion criteria were: aged from 20 to 45 years, not receiving treatment for premenstrual symptoms, regular menstrual cycles (25 to 38 days), no neuropsychiatric disorders, and had not taken drospirenone-containing oral contraceptive (OC)s, antidepressants, herbal medicine, probiotics, or antibiotics for 4 weeks before study entry. Drospirenone-containing OCs were excluded because, unlike other conventional OCs, they have a proven therapeutic effect on PMDs [30]. In addition, the inclusion criteria for the P group included at least one or more of the symptoms listed in the premenstrual symptoms questionnaire (PSQ) showing a moderate or higher level, and for the N group included no symptoms listed in the PSQ showing a moderate or higher level. In total, the microbiotas of 56 patients were analyzed. None of these patients had a history of inflammatory bowel disease, irritable bowel disease, malignancy, any gastrointestinal tract surgery, or diabetes, and all were OC non-users. According to the criteria for PMDs defined by the International Society of Premenstrual Disorder, there is no regulation on the number of premenstrual symptoms, but marked interference to the patient’s social life by the premenstrual symptoms is essential [5]. For an accurate diagnosis of PMD, it is recommended to keep a symptoms diary for two prospective periods, but this is not common in general gynecological practice in Japan. In this study, we further selected suspected cases of PMD (PMDs group) (n = 21) from the P group as those who experienced interference to their social life by the premenstrual symptoms. We also selected a control group (n = 22) from the N group who did not experience interference to their social life by premenstrual symptoms. The seven patients excluded from the N group had multiple mild premenstrual symptoms and mild interference to their social life due to these symptoms, as assessed by the PSQ.

Fig 1

Flow chart of the study design.

Abbreviation: PMDs, premenstrual disorders.

Questionnaire

The premenstrual symptoms questionnaire

In this study, we used the PSQ, developed in our previous study [31], to evaluate the severity of premenstrual symptoms. The PSQ has been found to have high reliability and validity [32].

The PSQ asks, “Within the last 3 months, have you experienced the following premenstrual symptoms starting during the week before menses and stopping a few days after the onset of menses?” The premenstrual symptoms listed are as follows: (i) depressed mood, (ii) anxiety or tension, (iii) tearfulness, (iv) anger or irritability, (v) decreased interest in work, home, or social activities, (vi) difficulty concentrating, (vii) fatigue or lack of energy, (viii) overeating or food cravings, (ix) insomnia or hypersomnia, (x) feeling overwhelmed, and (xi) physical symptoms such as tender breasts, feeling of bloating, headache, joint or muscle pain, or weight gain. These 11 symptoms are listed in the diagnostic criteria for PMDD in the Diagnostic and Statistical Manual of Mental Disorders (DSM). Furthermore, the PSQ asks whether these symptoms interfere with (a) work efficiency or productivity, or home responsibilities; (b) social activities; or (c) relationships with coworkers or family. The participants were asked to rate the severity of premenstrual symptoms and the interference with daily activities resulting from these symptoms as 1—Not at all, 2—Mild, 3—Moderate, or 4—Severe. The total scores of the PSQ were calculated as the sum of the 14 items. The PSQ total score was applied for the evaluation of premenstrual symptoms severity and ranges from 14 to 56.

Menstrual pain intensity

A numerical rating scale (NRS) was applied for the evaluation of menstrual abdominal pain intensity. On the NRS, study participants rate their severity of menstrual pain from 0 (no pain) to 10 (maximum pain you can imagine).

Evaluation of basic attributes

For each participant, we also collected information about age, parity, body weight, height, age at menarche, total sleep duration, smoking (yes or no), drinking (yes or no), regular exercise (yes or no), diarrhea or constipation. Body mass index (kg/m2) was calculated by dividing weight in kilograms by height in meters squared.

Measurement of inflammatory markers in blood samples

Blood samples were collected immediately after the PSQ and NRS assessment. Serum-separated samples were stored at −20°C until further analysis. Frozen samples were transferred to Filgen, Inc. (Nagoya, Japan) and CRP, soluble CD14 (sCD14), and lipopolysaccharide binding protein (LBP) were analyzed using the Human Magnetic Luminex Assay (R&D Systems, Inc., Minneapolis, MN, USA) according to the manufacturer’s protocols. All specimens were measured in triplicate and the average was obtained.

Analysis of the microbiota

All subjects were instructed to collect a stool sample at home. The participants scraped the surface of their stool with a swab after defecation and collected stool specimens in sampling tubes containing guanidine thiocyanate solution (Techno Suruga Laboratory Co. Ltd., Shizuoka, Japan). The DNA preservation solution ensures that the DNA of the stool specimen in the sampling tube is stable for at least one month at room temperature or at 4°C. The specimens were shipped by the subjects to our laboratory the same day using an express courier service, arriving no later than 48 hours after specimen collection. The collected specimens were sent to Techno Suruga Laboratory using a courier service at 4°C. The samples were stored at 4°C before DNA extraction. Extracted bacterial DNA was subjected to amplicon sequence analysis using the MiSeq system (Illumina, San Diego, CA, USA) by Techno Suruga Laboratory, as described previously [33]. Then, DNA was extracted using an automated DNA isolation system (GENE PREP STAR PI-480, Kurabo, Osaka, Japan), with 200 μL of distilled water being included as a negative control sample. The V3-V4 regions of Bacterial and Archaea 16S rRNA were amplified from the extracted DNA using the Pro341F/Pro805R primers and the dual-index method [33, 34], a negative control sample was also included, and the amplicons were visualized by electrophoresis. Barcoded amplicons were paired-end sequenced using a 2×284-bp cycle and the MiSeq system with MiSeq Reagent Kit version 3 (600 Cycle) chemistry. The quality of the paired-end sequencing reads was checked using the FASTX-Toolkit [35]. Paired-end sequencing reads were merged using the fastq-join program with default settings [36]. Only joined-reads that had a quality value score (QC) of ≥ 20 for more than 99% of the sequence were extracted using the FASTX-Toolkit [35]. Chimeric sequences were deleted with usearch6.1 [37, 38]. Bacterial and Archaea species identification from sequences was performed using the Metagenome@KIN Ver 2.2.1 analysis software (World Fusion, Japan) and the TechnoSuruga Lab Microbial Identification database DB-BA 13.0 (TechnoSuruga Laboratory, Japan) with homology of ≥ 97% [39]. The relative abundance of each bacterium at the phylum and genus level was calculated.

The 16S rRNA data were also analyzed with Quantitative Insights into Microbial Ecology (QIIME) 2.0 ver. 2020.6 [40]. Quality filtering and chimeric sequences were filtered using DADA2 (Divisive Amplicon Denoising Algorithm 2) denoise-single plugin ver. 2017.6.0 with default option [41]. Taxonomy was assigned using Greengenes database ver. 13.8 based on an average percent identity of 99% [42] by training a Naive Bayes classifier using the q2-feature-classifier plugin. To analyze beta diversity, weighted unifrac distance metrics were used. We used principal coordinates analysis (PCoA) to show the pattern of differences in the PMDs group and the control group. Alpha diversity was calculated by the Chao 1 [43], Shannon [44], and Simpson indices [45]. To further investigate the differences in abundance between the PMDs group and control group at the genus level, linear discriminant effect size analysis (LEfSe) was performed through the Huttenhower Lab Galaxy Server [46]. LEfSe was performed under the following conditions: the alfa value for the Kruskal–Wallis test was 0.05 and the threshold for the logarithmic linear discriminant analysis (LDA) score for a discriminative feature was 2.0.

Statistical analysis

For continuous variables, normally distributed data were expressed as means and standard deviations and were analyzed by the Student’s t-test, while non-normally distributed data were expressed as medians and interquartile ranges and were analyzed by the Wilcoxon signed-rank test. For the Student’s t-test, effect size was measured using r, and r was calculated by the following formula (r = ). For the Wilcoxon signed-rank test, effect size was measured using r, and r was calculated by the following formula (r = ). For categorical variables, proportions were calculated and analyzed by Fisher’s exact test. Effect size was measured using Cramer’s V. The effect sizes of 0.10, 0.30, and 0.50 were judged as small, medium, and large, respectively [47].

For the relative abundance analysis of gut microbiota, multiple comparisons were adjusted using false discovery rate (FDR) correction. An FDR-adjusted P value (q value) < 0.20 was set as the cut-off [48].

Correlations between gut microbiota abundance and PSQ total score were analyzed using Spearman’s rank correlation coefficient. Multiple regression analysis was used to explore the association between the microbiota and the PSQ total score. Variables that were predictive at a P value less than 0.20 were introduced into the stepwise model.

Statistical analyses were performed using JMP Pro 15.1.0 (SAS, Cary, NC, USA), except for the relative abundance analysis of the gut microbiota, for which SAS 9.4 (SAS) was used. Statistical significance was set at P < 0.05 (for two-tailed tests).

Results

The characteristics of the study population are presented in Table 1.

Table 1

Characteristics of the study participants.

Characteristic	Total (n = 56)	PMDs (n = 21)	Control (n = 22)	P
Age (years), median (IQR)	27.5 (23.0–35.0)	26.0 (23.0–31.0)	26.0 (22.8–37.5)	0.394a (r = 0.130)
Parity, median (IQR)	0 (0–1.8)	0 (0–2.0)	0 (0–0)	0.073a (r = 0.273)
Age at menarche (years), median (IQR)	12.0 (11.0–13.0)	12.0 (11.0–13.0)	12.5 (11.0–14.0)	0.397a (r = 0.129)
BMI (kg/m2), median (IQR)	21.7 (19.3–23.1)	21.2 (19.1–22.7)	20.4 (19.0–23.4)	0.799a (r = 0.039)
Menstrual pain intensity, median (IQR)	4.5 (3.0–7.0)	8.0 (5.5–8.0)	3.0 (1.0–4.0)	<0.0001a (r = 0.656)
Total sleep duration (hours), median (IQR)	6.0 (6.0–7.0)	6.0 (6.0–7.0)	6.3 (6.0–7.0)	0.207a (r = 0.192)
Smoker, n (%)	9.0 (16.1)	1.0 (4.8)	6.0 (27.3)	0.095b (V = 0.305)
Drinker, n (%)	19.0 (33.9)	6.0 (28.6)	9.0 (40.9)	0.526b (V = 0.129)
Regular exercise, n (%)	7.0 (12.7)	3.0 (15.0)	1.0 (4.6)	0.333b (V = 0.178)
Diarrhea or constipation, n (%)	23.0 (41.1)	10.0 (47.6)	7.0 (31.8)	0.358b (V = 0.162)
PSQ total score, median (IQR)	22.0 (18.0–30.0)	32.0 (28.5–40.5)	17.0 (15.8–18.3)	<0.0001a(r = 0.844)

Abbreviations: PMDs, premenstrual disorders; IQR, interquartile range; BMI, body mass index; SD, standard deviation; PSQ, premenstrual symptoms questionnaire; V, Cramer’s V

a Wilcoxon signed-rank test

b Fisher’s exact test

The severity of menstrual pain was stronger and the PSQ total score was higher in the PMDs group than in the control group (P < 0.0001).

The differences in endotoxin biomarkers between the two groups are presented in Table 2.

Table 2

Comparison of the endotoxin biomarkers present in the PMDs group and the control group.

Characteristic	PMDs (n = 21)	Control (n = 22)	P
CRP (ng/ml), median (IQR)	308.2 (117.0–477.6)	242.3 (61.1–980.6)	0.743a (r = 0.050)
s-CD14 (ng/ml), mean (SD)	918.2 (37.7)	912.7 (36.8)	0.917b (r = 0.016)
LBP (ng/ml), median (IQR)	5859.6 (5213.7–6579.0)	5404.2 (4506.6–7312.7)	0.536a (r = 0.093)

Abbreviations: PMDs, premenstrual disorders; CRP, C reactive protein; IQR, interquartile range; s-CD14, soluble cluster of differentiation 14; SD, standard deviation; LBP, lipopolysaccharide binding protein

a Wilcoxon signed-rank test

b Student’s t-test

There was no significant difference between the expression levels of endotoxin biomarkers in the PMDs group and the control group.

Next, we analyzed alpha and beta diversity in the two groups (Fig 2). Regarding alpha diversity, there were no significant differences between the PMDs group and the control group as determined by the Chao 1, Shannon, and Simpson indices (P = 0.430, 0.423, and 0.308, respectively). Regarding beta diversity, as analyzed by PCoA, a significant difference was detected between the PMDs group and the control group (R = 0.062, P = 0.027).

Fig 2

Principal coordinates analysis plot comparing sample distribution between the PMDs group and the control group.

Each blot shows data for alpha diversity, the Chao 1 index (A), the Shannon index (B), and the Simpson index (C). Abbreviations: C, control group; P, premenstrual disorders group.

The relative abundance of organisms in the gut microbiota was compared between the PMDs group and the control group (Fig 3). At the phylum level, the PMDs group possessed fewer Bacteroidetes than the control group (P = 0.015, q = 0.136) (Fig 3A). We further analyzed the Firmicutes/Bacteroidetes (F/B) ratio, but this did not differ significantly between the two groups (P = 0.111). At the genus level, the PMDs group had a lower prevalence of Butyricicoccus, Extibacter, Megasphaera, Parabacteroides, and “Not determined” (P = 0.037, 0.018, 0.028, 0.039 and 0.033, respectively), and a higher prevalence of Anaerotaenia (P = 0.017) than the control group; however, after FDR correction, this significance was lost (Fig 3B).

Fig 3

Comparison of the gut microbiotas between the PMDs group and the control group.

The relative abundance of each taxon in the gut microbiota was compared. (A) At the phylum level, only Bacteroidetes was significantly less abundant in the PMDs group than in the control group. (B) The abundance of genus-level bacteria was significantly different between the PMDs group and the control group. The Wilcoxon signed-rank test was used to compare differences between the two groups (P < 0.05). Abbreviations: PMDs, premenstrual disorders; P, premenstrual disorders group; C, control group.

Furthermore, we analyzed the characteristic gut microbiota differences in the PMDs group and the control group by LEfSe (Fig 4A and 4B). At the genus level, Anaerotaenia was enriched in the PMDs group, whereas Extibacter, Butyricicoccus, “Not determined”, Megasphaera, and Parabacteroides were enriched in the control group (Fig 4B).

Fig 4

Linear discriminant effect size analysis to distinguish the differential microbiota between the PMDs group and control group.

(A) Cladogram showing the most differentially abundant taxa between the PMDs group (P) and control group (C). Taxa enriched in the PMDs group are indicated in green and taxa enriched in the control group are indicated in red. The brightness of each dot is proportional to the respective effect size. (B) Comparison of the LDA scores between the P and C groups at the genus level. Abbreviations: LDA, logarithmic linear discriminant analysis; PMDs, premenstrual disorders; P, premenstrual disorders group; C, control group.

Next, we analyzed the association between the abundance of organisms in the gut microbiota and the severity of premenstrual symptoms (Table 3).

Table 3

Correlation analysis between the gut microbiota abundance and the PSQ total score (n = 56).

	R	P
Anaerotaenia	0.292	0.029
Extibacter	−0.410	0.002
Parabacteroides	−0.342	0.010

Abbreviations: PSQ, premenstrual symptoms questionnaire

At the genus level, the PSQ total score was positively associated with Anaerotaenia, and negatively associated with Extibacter and Parabacteroides. Multiple regression analysis showed that the PSQ total score was negatively associated with Parabacteroides and Megasphaera (Table 4).

Table 4

Multiple regression analysis calculating the associations between the microbiota and the PSQ total score (n = 56).

	β	95% CI	P	Standardized β	VIF
Blautia	0.36	−0.07 to 0.79	0.10	0.23	1.24
Faecalibacterium	−0.33	−0.82 to 0.17	0.19	−0.17	1.07
Parabacteroides	−1.31	−2.43 to -0.18	0.02	−0.30	1.14
Ruminococcus	1.15	−0.07 to 2.36	0.06	0.25	1.27
g_Lachnospiraceae bacterium KNHs209_incertae_sedis	−2.65	−5.36 to 0.06	0.06	−0.24	1.10
Megasphaera	−1.56	−2.89 to -0.24	0.02	−0.31	1.19

R = 0.29

Abbreviations: PSQ, premenstrual symptoms questionnaire; β, regression coefficient; CI, confidence interval; VIF, variance inflation factor

Variance inflation factor analysis showed that multicollinearity was not present for premenstrual symptoms in this model.

Discussion

To our knowledge, this is the first report to investigate the association between the gut microbiota and premenstrual symptoms. By amplicon sequencing, we detected a difference in the gut microbiota between the PMDs group and the control group. Furthermore, we identified several organisms within the gut microbiota that were significantly associated with the severity of premenstrual symptoms.

In the case of MDD, increased bacterial translocation has been suggested to play a role in the inflammatory pathophysiology [24]. In this study, rather than directly assessing bacterial translocation by analyzing the presence of gut bacteria in the blood, we instead indirectly assessed bacterial translocation by measuring the levels of inflammatory factors CRP, LBP, and sCD14. LBP and sCD14 are produced in response to bacterial translocation and are proposed markers of endotoxemia [24]. Given the exploratory nature of this study, we chose the indirect method of evaluation as a simpler method for this initial investigation. According to our data, these indicators of inflammation did not show any correlation with PMDs. MDD and PMDs are closely related in terms of their clinical symptoms. However, the duration of symptoms is distinct between these two disorders, with MDD symptoms being permanent and PMD symptoms being temporary. This may explain the difference in the degree of inflammation between MDD and PMDs.

There was no significant difference in alpha diversity between the PMDs group and the control group. According to data from meta-analysis of MDD patients, there was no significant difference in alpha diversity between MDD patients and controls [18]. In other diseases, such as inflammatory bowel disease, obesity, and metabolic diseases, decreased microbiota diversity was suggested to be associated with the development of these diseases [49, 50]. This may explain some of the common pathology of MDD and PMDs.

At the phylum level, Bacteroidetes was less abundant in the PMDs group than in the control group. Low Bacteroidetes levels have been reported to be associated with obesity and MDD [51, 52]. Furthermore, obesity was reported to be a risk factor for PMDs and MDD [53, 54]. Considering that, in this study, no significant difference was found in BMI between the PMDs group and control group, obesity did not seem to be a confounding factor.

At the genus level, our data indicated that, in general, decreased levels of Butyricicoccus, Extibacter, Megasphaera, and Parabacteroides were associated with PMDs. These gut microorganisms were different from those reported to be decreased in MDD patients compared with non-MDD patients [18].

Among the gut microbiota, Butyricicoccus is a butyrate-producing beneficial bacterium and Megasphaera metabolizes lactate to butyrate [55]. In animal models, butyrate-treated mice possessed an increased amount of brain-derived neurotrophic factor (BDNF), which is essential for nerve cell growth and has been linked to antidepressant effects [56]. Decreased levels of butyrate-producing bacteria, such as Butyricicoccus and Megasphaera, may be involved in the pathology of PMDs.

Decreased levels of Butyricicoccus have been reported in postpartum depressive disorder (PPD) [57]. PPD is a unique subtype of MDD, for which the precise pathogenesis remains unknown. During pregnancy, women possess high levels of estrogen and progesterone because of the presence of the placenta and fetus. Dramatic hormonal fluctuations occur after delivery, which is thought to be one of the causes of PPD [58]. The hormonal fluctuations observed with PPD are the same as for PMDs, and premenstrual symptoms have been proposed as a risk factor for PPD [59]. Decreased levels of Butyricicoccus may explain some of the common pathology of PPD and PMDs.

Parabacteroides is reported to be beneficial in protecting against multiple sclerosis [60], seizures [61], metabolic dysfunctions [62], and colon tumors [63]. Regarding the gut microbiota–brain axis, Parabacteroides was reported to be a GABA-producer according to the results of GABA-dependent co-culture assays and in silico analyses [64]. In a seizure mouse model, the anti-seizure effects of a ketogenic diet were analyzed in the gut microbiota [61]. In this study, Parabacteroides was shown to modulate brain GABA levels. Because impaired GABA function is one of the possible causes of PMDs [7], decreased levels of Parabacteroides may be involved in PMD pathology.

Our study had several limitations. The main limitation was that the study was cross-sectional in design. It was therefore impossible to determine causality between premenstrual symptoms and the gut microbiota changes. To clarify the causality, further longitudinal studies are required. Second, our study had a small sample size; however, we believe this was appropriate for a pilot study. Third, we selected the participants from outpatients seeking care. This is a highly select population, which could have led to data bias. Fourth, we selected the PMDs group without prospective daily charting over two consecutive symptomatic cycles, which is recommended by the DSM [30]. However, prospective daily charting is difficult in clinical settings. According to a report from the USA, only 11.5% of physicians reported routinely monitoring two consecutive symptomatic cycles [65]. Fifth, the timing of blood and stool sample collection was performed without considering the menstrual cycle. There is a possibility that the gut microbiota may fluctuate depending on the stage of the menstrual cycle, and sampling at a consistent stage of the menstrual cycle may be necessary in future investigations. Sixth, the study was conducted only in Japan, which might limit the generalization of the findings to the other countries, especially western countries. Finally, the significant difference in the relative abundance of the gut microbiota between the PMDs group and the control group was lost after FDR correction. Considering that the results of LEfSe showed the same pattern of differences in abundance between the PMDs group and the control group, these differences in the gut microbiota between the two groups are likely to be meaningful. Further studies considering these factors are needed to confirm the reliability of these findings. Consecutive prospective evaluations using the Daily Record of Severity of Problem (DRSP) are recommended for accurate assessment of premenstrual symptoms [30], and it is possible to use the Japanese version of the DRSP, for which the validity and reliability have been confirmed [66]. Using DRSP assessment in a large-scale prospective study of the general public, and performing blood and stool collection at specific times during the follicular and luteal phases, respectively, would be expected to provide more definitive results.

Despite these limitations, differences in the gut microbiota between PMD patients and healthy individuals may be applied as biomarkers for diagnosis in the future.

Conclusions

The present study showed that gut microbial properties were associated with premenstrual symptoms. Decreased levels of Parabacteroides and Megasphaera are a characteristic feature of PMD patients and are negatively associated with the severity of premenstrual symptoms.

Data for Table 1.

(XLSX)

Click here for additional data file.

Data for Table 2.

(XLSX)

Click here for additional data file.

Phylum-level operational taxonomic units of the fecal microbiota.

(XLSX)

Click here for additional data file.

Genus-level operational taxonomic units of the fecal microbiota.

(XLSX)

Click here for additional data file.

2 Nov 2021

rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Jonathan Jacobs

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

4.  We noticed you have some minor occurrence of overlapping text with the following previous publication, which needs to be addressed:

- https://www.dovepress.com/psychometric-testing-of-the-premenstrual-symptoms-questionnaire-and-th-peer-reviewed-fulltext-article-IJWH

The text that needs to be addressed involves the first paragraph of the Introduction.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Takeda et al studied fecal microbiome using 16s rRNA sequencing in premenstrual disorder (PMD) patients compared to control women. They found differences in the composition of bacteria in women with PMD compared to control women. This is an interesting study and has attempted to address an important question. However, there are several issues pointed below, that need to be addressed.

1. The sample sizes in abstract and the main text are different (27 PMS and 29 controls in abstract vs 30 each in the main text). Additionally, it is unclear if the entire group was studied or the PMD subgroup was studied for microbiome-related analyses. The authors suggest that the subset of control subjects studied (N=22) did not have PMS or social disturbance. It is not clear what the authors mean by “social disturbance”. If it is one of the features of PMS, why were 8 controls (30-22) subjects included with “social disturbance”? Is there a microbial analysis performed in the larger group as well as in the smaller group? If not, the actual sample size of the study would be 22 and 21. It would help if the terms PMS, PMDD and PMD be described initially in the manuscript and only those abbreviations be used when referring the dataset consistently. For example, line 90, the recruitment was for “patients who wished to receive treatment”. If these were PMS, then it should be stated something on these lines - “Thirty patients with PMS were recruited …” and state how many had PMD or PMDD.

2. The entire Methods section of the abstract needs to be reworded for better clarity and correctness (for example, “complaining premenstrual symptoms”, which can be reworded as “experiencing ..”).

3. Aims stated on line 83 do not sufficiently capture the analysis and should be clearly stated.

4. If a correlation was tested for all genera, multiple testing corrections need to be applied and significance threshold needs to be implemented at an appropriate FDR cutoff (typically 5 or 10%) instead of a p-value cutoff.

5. Language and grammar need to be improved and space between words need to be fixed at several places in the abstract and the main manuscript. Some places are listed below

a. Line 139: should be “triplicate”.

b. Statistical analysis: “abnormally distributed” should be restated as “non-normally”; “Medium” should be “median”.

c. Line 248: Needs to be reworded.

d. Line 280: “High gonadal condition” is unclear.

6. 252: How did the authors measure “bacterial translocation”? The entire paragraph describing inflammation and bacterial translocation is unclear.

7. Asterisk usually refers to statistical significance. Please use another symbol in Table 1 to describe the tests used.

8. As per Figure 3B, control group had very low abundance of Anaerotaenia and Extibacter and PMD group had very low abundance of Megasphaera. It would be better to include points/jitter. Looks like it may be driven by outliers.

Reviewer #2: This paper presents data from a cross-sectional study on relationship between gut microbiota and premenstrual symptoms. The purpose of the study is innovative and interesting; however, multiple design and methodological flaws dampen my enthusiasm for the study. My detailed comments and suggestions are below.

Abstract:

1. The scope of the study can be clearer. Did the authors focus broadly on premenstrual symptoms? Premenstrual syndrome? or PMDD?

2. The term “social disturbance” is confusing. It sounds like social unrest. I believe the author mean disruption or interference of social life.

3. What type of microbial diversity did you refer to in the abstract?

4. It would be helpful to report P values and effect sizes when possible.

Data availability:

5. It is unclear why deidentified data underlying the findings cannot be made fully available. The openly sharing of deidentified data is important to safeguard the reproducibility of the study findings.

Background:

6. Again, the focus of the study needs to be clearer. Premenstrual syndrome? PMDD? Both PMS and PMDD? Premenstrual symptoms in general?

7. Line 60: Certain lifestyle factors have been found to be associated with premenstrual symptoms. However, it is imprecise to say that poor lifestyle habits cause the symptoms.

8. Line 4: What are the characteristics of the “systematic inflammation disorder”?

9. The rationale for selected blood biomarkers can be clearer. There are many measures for intestinal permeability and bacteria translocation. Have markers selected by the authors been linked to premenstrual symptoms in previous research?

10. Pre-specified hypotheses need to be clearly stated.

Methods:

11. Many possible comorbidities that are known to affect gut microbiome were neither considered in the design nor the the analysis (inflammatory bowel diseases, irritable bowel syndrome, diabetes). This is a major limitation of the study. Please acknowledge this in the discussion.

12. Similarly, is there information available regarding psycholocial comorbidities among participants?

13. The sample were selected patients seeking care at the outpatient ob/gyn clinic. This can be a highly selected population. Please acknowledge this as a limitation.

14. The antibiotic usage before sample collection was not clear.

15. What was the rationale for sample size?

16. Case and control need to be better defined. What is the rationale for case and control selection?

17. The case definition seems quite inconsistent with the DSM-3 criteria. The DSM criteria require to have at least 5 symptoms for PMDD diagnosis. In the study, the presence of one moderate or severe symptom will qualify the participants as cases. Also, the symptoms within the last 3 months can be different from most menstrual cycles (part of DSM-definition). These are major limitations of the study.

18. For menstrual pain intensity, which type of menstrual pain was assessed? Abdominal? Menstrual headache? Both?

19. When were stool and blood samples collected? Was menstrual cycle stage controlled? Sex hormones can potentially affect gut microbiome profile.

20. It is unclear if rectal swab or stool was collected. What was the length of time from sample collection to sample receipt? How was the sample stored?

21. For microbiome assays, were positive and negative controls used? If so, please describe. Also, what quality control measures were used?

22. For data analysis: What correction methods were used for multiple comparisons? If no correction method was used, please acknowledge this as a limitation.

Results:

23. For effect size measures from the LEfSE analysis, it would be helpful to show the plots in the results.

Discussion:

24. Please discuss and potential confounders (e.g., comorbid gastrointestinal and psychological factors) that may influence/confound the study results.

25. There are many different measures of bacterial translocation in the literature. Please comment the quality of the bacterial translocation measure used in the study. This will help readers interpret the findings related to blood biomarker measures.

26. Please acknowledge several limitations noted in my comments of the methods section.

27. The implications for future research can be clearer. What further research is needed to further investigate the mechanisms of gut microbiome- premenstrual symptom association?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Swapna Mahurkar-Joshi

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

16 Dec 2021

Professor Emily Chenette

Editor-in-Chief

PLOS ONE

December 16, 2021

Dear Professor Chenette,

Thank you for reconsidering our manuscript entitled “Characteristics of the gut microbiota in women with premenstrual symptoms: A cross-sectional study” as an Original Research article for publication in PLOS ONE. We have revised our manuscript in response to the many helpful comments of the reviewers, as shown below.

Academic Editor’s comments

1) There are extensive problems with the clarity of the writing (many specific examples are provided by the reviewers). I recommend the use of an English language writing service to assist with revisions.

Response (Re): Our revised manuscript has been reviewed by a native English-speaking expert editor.

2) More precise description of the study population is required (e.g. distinction between PMS vs. PMDD, clarification of "social disturbance" as a factor for selecting the two study groups, which subjects were included in each of the microbiome analyses that are shown)

Re: We have added an explanation about the difference between PMS, PMDD, and PMDs on page 4, lines 57–60. We further added the criteria for PMDs and explained the importance of “social disturbance” (or “interference with social life”) in these criteria, as described on page 6, lines 115–118.

3) Statistical analyses need to be corrected for multiple hypothesis testing (and if significance is lost after FDR correction, this needs to be addressed as a limitation)

Re: We applied FDR correction and added the data on the q values to Figure 3. We also added an explanation of this to the Materials and Methods section on page 11, lines 218–220, and added a new reference (47). In the original version, the JMP software was used for statistical analysis, but since FDR correction was not possible, the analysis was redone using the SAS software. As a result of this revision, the P value for the Wilcoxon signed-rank test changed slightly, so the results were revised in Figure 3 and on page 14, lines 274–275. In conducting this additional analysis, Keiko Yamada was added as a co-author. The difference in relative abundance of the gut microbiota between the PMDs group and the control group was lost after FDR correction and we have added an explanation of this on page 3, lines 40–41 and page 14, lines 275–276. Furthermore, we have added this to the limitations section on page 21, lines 391–396.

4) Deidentified data needs to be made available per PLOS ONE policy

Re: The data set supporting the results cannot be made accessible to the public because it contains potentially sensitive patient information. Based on the ethical guidelines in Japan, the ethics committee of Kindai University has imposed restrictions on the dissemination of the data in this study. The use of this data set requires approval by the ethics committee of Kindai University. In the case of a data access request, please contact the corresponding author: Takashi Takeda (take@med.kindai.ac.jp).

Journal Requirements:

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

Re: We have followed the PLOS ONE style as indicated.

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/ plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Re: The data set supporting the results cannot be made accessible to the public because it contains potentially sensitive patient information. Based on the ethical guidelines in Japan, the ethics committee of Kindai University has imposed restrictions on the dissemination of the data in this study. The use of this data set requires approval by the ethics committee of Kindai University. In the case of a data access request, please contact the corresponding author: Takashi Takeda (take@med.kindai.ac.jp).

3. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

Re: We have moved the ethics statement to the Materials and Methods section as indicated.

4. We noticed you have some minor occurrence of overlapping text with the following previous publication, which needs to be addressed:

The text that needs to be addressed involves the first paragraph of the Introduction.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

Re: According to this comment, we have removed the overlapping text and correctly cited all sources on page 4, lines 54–64.

Reviewer: 1

1. The sample sizes in abstract and the main text are different (27 PMS and 29 controls in abstract vs 30 each in the main text). Additionally, it is unclear if the entire group was studied or the PMD subgroup was studied for microbiome- related analyses. The authors suggest that the subset of control subjects studied (N=22) did not have PMS or social disturbance. It is not clear what the authors mean by “social disturbance”. If it is one of the features of PMS, why were 8 controls (30-22) subjects included with “social disturbance”? Is there a microbial analysis performed in the larger group as well as in the smaller group? If not, the actual sample size of the study would be 22 and 21. It would help if the terms PMS, PMDD and PMD be described initially in the manuscript and only those abbreviations be used when referring the dataset consistently. For example, line 90, the recruitment was for “patients who wished to receive treatment”. If these were PMS, then it should be stated something on these lines - “Thirty patients with PMS were recruited …” and state how many had PMD or PMDD.

Re: As pointed out by the reviewers, the definition of subjects was unclear in the original version. In the present study, there was no diagnosis of PMS, PMDD, or PMD in the strict sense of the word, only patients who came to the clinic for treatment of premenstrual symptoms. For an accurate diagnosis of PMS, PMDD, or PMD, it is necessary to keep a symptom diary for two prospective periods, but this is not common in general gynecological practice in Japan. We have added an explanation of this on page 6, lines 118–120. Groups P and N in the text were those who meet the inclusion and exclusion criteria, and the description of 30 women in each group was incorrect and has been revised to 27 women in P group and 29 women in N group as described on page 6, lines 104–106. In line with this revision, the description in Figure 1 has also been changed. According to the criteria for PMDs by the International Society of Premenstrual Disorder, there is no regulation on the number of premenstrual symptoms, but marked interference with the patient’s social life by the premenstrual symptoms is essential. According to this background, we selected suspected cases of PMD (the PMDs group) from the P group based on interference with their social life. We have added this explanation on page 6, lines 115–121, and page 7, line 122–124. We also corrected the expression “PMS” to “PMDs” on page 3, line 43.

2. The entire Methods section of the abstract needs to be reworded for better clarity and correctness (for example, “complaining premenstrual symptoms”, which can be reworded as “experiencing ..”).

Re: According to this comment, we reworded “complaining premenstrual symptoms” to “experiencing premenstrual symptoms”, and “social disturbance” to “interference to their social life” on page 2, lines 27–29.

5. Language and grammar need to be improved and space between words need to be fixed at several places in the abstract and the main manuscript. Some places are listed below

a. Line 139: should be “triplicate”.

Re: According to this comment, “triple cate” has been corrected to “triplicate”. Our revised manuscript has also been reviewed by a native English-speaking expert editor.

b. Statistical analysis: “abnormally distributed” should be restated as “non-normally”; “Medium” should be “median”.

Re: According to this comment, we reworded “abnormally distributed” to “non-normally”, and “Medium” to “median”.

c. Line 248: Needs to be reworded.

Re: According to this comment, we added “amplicon sequence analysis” to the explanation of next generation sequencing. In connection with this change, the description of next generation sequencing was revised on page 4, line 70, page 9, line 174 and page 17, line 323.

d. Line 280: “High gonadal condition” is unclear.

Re: According to this comment, we reworded “high gonadal condition” to “high levels of estrogen and progesterone” on page 19, lines 362–363.

6. 252: How did the authors measure “bacterial translocation”? The entire paragraph describing inflammation and bacterial translocation is unclear.

Re: Rather than directly assessing bacterial translocation by analyzing the presence of gut bacteria in the blood, we instead indirectly assessed bacterial translocation by measuring the levels of inflammatory factors CRP, LBP, and sCD14. LBP and sCD14 are produced in response to bacterial translocation and are proposed markers of endotoxemia. We added this explanation on page 17, lines 328–333.

7. Asterisk usually refers to statistical significance. Please use another symbol in Table 1 to describe the tests used.

Re: We changed the asterisk to alphabetic characters in Table 1 and Table 2.

8. As per Figure 3B, control group had very low abundance of Anaerotaenia and Extibacter and PMD group had very low abundance of Megasphaera. It would be better to include points/jitter. Looks like it may be driven by outliers.

Re: We changed the graph display method to include points/jitter in Figure 3.

Reviewer: 2

Abstract:

1. The scope of the study can be clearer. Did the authors focus broadly on premenstrual symptoms? Premenstrual syndrome? or PMDD?

Re: The scope of the study was PMDs and premenstrual symptoms. We have clarified the text in the Abstract accordingly on page 2, lines 21–22.

2. The term “social disturbance” is confusing. It sounds like social unrest. I believe the author mean disruption or interference of social life.

Re: According to this comment, we have changed “social disturbance” to “interference to their social life” on page 2, lines 28 and 29 and page 7, line 122–124.

3. What type of microbial diversity did you refer to in the abstract?

Re: In the Abstract, we described alpha diversity and beta diversity. We have added this detail on page 2, lines 36.

4. It would be helpful to report P values and effect sizes when possible.

Re: We analyzed the effect sizes and have added these data to Tables 1 and 2. We also added an explanation of this analysis on page 10, lines 212–213 and page 11, lines 214–217.

Data availability:

5. It is unclear why deidentified data underlying the findings cannot be made fully available. The openly sharing of deidentified data is important to safeguard the reproducibility of the study findings.

Re: The data set supporting the results cannot be made accessible to the public because it contains potentially sensitive patient information. Based on the ethical guidelines in Japan, the ethics committee of Kindai University has imposed restrictions on the dissemination of the data in this study. The use of this data set requires approval by the ethics committee of Kindai University. In the case of a data access request, please contact the corresponding author: Takashi Takeda (take@med.kindai.ac.jp).

Background:

6. Again, the focus of the study needs to be clearer. Premenstrual syndrome? PMDD? Both PMS and PMDD? Premenstrual symptoms in general?

Re: The scope of the study was PMDs and premenstrual symptoms. We have clarified this point on page 5, lines 89–92.

7. Line 60: Certain lifestyle factors have been found to be associated with premenstrual symptoms. However, it is imprecise to say that poor lifestyle habits cause the symptoms.

Re: We agree and have modified the text accordingly on page 4, line 62–63.

8. Line 4: What are the characteristics of the “systematic inflammation disorder”?

Re: This expression was inaccurate and has been corrected to “low-grade inflammation” on page 5, line 76.

9. The rationale for selected blood biomarkers can be clearer. There are many measures for intestinal permeability and bacteria translocation. Have markers selected by the authors been linked to premenstrual symptoms in previous research?

Re: Although LBP and sCD14 have not been linked to premenstrual symptoms previously, they have been studied in relation to depressive symptoms. We have added the relevant references for this (25, 26) and an explanation on page 5, lines 82–83.

10. Pre-specified hypotheses need to be clearly stated.

Re: We have added the hypotheses and aims of this study at the end of the Introduction on page 5, lines 87–92.

Methods:

11. Many possible comorbidities that are known to affect gut microbiome were neither considered in the design nor the the analysis (inflammatory bowel diseases, irritable bowel syndrome, diabetes). This is a major limitation of the study. Please acknowledge this in the discussion.

Re: These diseases were not considered in this study, and patients with these diseases were excluded. We have acknowledged this on page 6, lines 113–115.

12. Similarly, is there information available regarding psycholocial comorbidities among participants?

Re: We excluded any patients with neuropsychiatric disorders from the study. We have added this information to the inclusion criteria on page 6, line 108.

13. The sample were selected patients seeking care at the outpatient ob/gyn clinic. This can be a highly selected population. Please acknowledge this as a limitation.

Re: We have added this as a limitation on page 20, lines 380–382.

14. The antibiotic usage before sample collection was not clear.

Re: We specify the absence of antibiotic usage in the inclusion criteria on page 6, line 109.

15. What was the rationale for sample size?

Re: Since there are no reports on the relationship between premenstrual symptoms and gut microbiota, we conducted a pilot study. Normally, we consider 20 to 30 cases to be appropriate for such an exploratory study; however, we have mentioned this as a limitation of the study on page 20, lines 379–380.

16. Case and control need to be better defined. What is the rationale for case and control selection?

Re: There were two parts to our study. The first was to compare suspected PMD cases with control cases, and the second was to examine the relationship between the severity of premenstrual symptoms and the gut microbiota. This is clearly stated as the aim of the study in the last part of the Introduction on page 5, lines 89–92. In addition, the diagnosis of PMDs has been added to the Materials and Methods section on page 6, lines 115–120.

17. The case definition seems quite inconsistent with the DSM-3 criteria. The DSM criteria require to have at least 5 symptoms for PMDD diagnosis. In the study, the presence of one moderate or severe symptom will qualify the participants as cases. Also, the symptoms within the last 3 months can be different from most menstrual cycles (part of DSM-definition). These are major limitations of the study.

Re: The current study targets PMDs and not PMS or PMDD. The differences between these disorders have been clarified in the Introduction on page 4, lines 57–60. In addition, the criteria for the diagnosis of PMD have been added to the Materials and Methods section on page 6, lines 115–120.

18. For menstrual pain intensity, which type of menstrual pain was assessed? Abdominal? Menstrual headache? Both?

Re: Menstrual pain usually refers to abdominal pain. We have clarified this on page 8, line 151.

19. When were stool and blood samples collected? Was menstrual cycle stage controlled? Sex hormones can potentially affect gut microbiome profile.

Re: The timings of blood and stool collection were conducted without considering the menstrual cycle. We have added this as a limitation of the study on page 20, lines 386–389.

20. It is unclear if rectal swab or stool was collected. What was the length of time from sample collection to sample receipt? How was the sample stored?

Re: Stool samples were collected by swabbing immediately after defecation. We have added this information on page 9, lines 170–171, and line 173.

21. For microbiome assays, were positive and negative controls used? If so, please describe. Also, what quality control measures were used?

Re: Negative controls were included as described on page 9, lines 177–178, and lines 180. We also describe the quality control measures on page 9, lines 183–184.

22. For data analysis: What correction methods were used for multiple comparisons? If no correction method was used, please acknowledge this as a limitation.

Re: We applied FDR correction and have added the q-value data to Figure 3, along with an explanation in the Materials and Methods section on page 11, lines 218–220, and a new reference (47). In the original version, the JMP software was used for statistical analysis, but since FDR correction was not possible, the analysis was redone using the SAS software. As a result of this revision, the P value resulting from the Wilcoxon signed-rank test changed slightly, so the results were revised in Figure 3 and on page 14, lines 274–275. To conduct this additional analysis, Keiko Yamada was added as a co-author. The difference in the relative abundance of the gut microbiota between the PMDs group and the control group was lost after FDR correction and we have added this information on page 3, lines 40–41 and page 14, lines 275–276. We have also added this as a limitation on page 21, lines 391–396.

23. For effect size measures from the LEfSE analysis, it would be helpful to show the plots in the results.

Re: We performed LEfSE through the Huttenhower Lab Galaxy Server, but we were unable to use the plots for the results from this server.

Discussion:

24.Please discuss and potential confounders (e.g., comorbid gastrointestinal and psychological factors) that may influence/confound the study results.

Re: As stated in our response to comments 11 and 12, these potential confounders were not considered in our study. Furthermore, the timings of blood and stool collection did not take the menstrual cycle into consideration and this has been added as a limitation on page 20, lines 386–389.

25.There are many different measures of bacterial translocation in the literature. Please comment the quality of the bacterial translocation measure used in the study. This will help readers interpret the findings related to blood biomarker measures.

Re: We have added an explanation of the blood biomarkers used on page 17, lines 328–333.

26.Please acknowledge several limitations noted in my comments of the methods section.

Re: We have added a comprehensive list of the limitations of our study on page 20, lines 380–382, page 20, lines 386–389 and page 21, lines 391–396.

27.The implications for future research can be clearer. What further research is needed to further investigate the mechanisms of gut microbiome- premenstrual symptom association?

Re: We have added the implications in terms of future research to the Discussion section on page 21, lines 397–403.

Submitted filename: Response to Reviewrs.docx

Click here for additional data file.

10 Mar 2022

24 Mar 2022

Professor Emily Chenette

Editor-in-Chief

PLOS ONE

March 24, 2022

Dear Professor Chenette,

Thank you for reconsidering our manuscript entitled “Characteristics of the gut microbiota in women with premenstrual symptoms: A cross-sectional study” as an Original Research article for publication in PLOS ONE. We have revised our manuscript in response to the many helpful comments of the reviewers, as shown below.

Additional Editor Comments:

The primary remaining issue is data availability. Please see the policy of PLOS ONE: https://journals.plos.org/plosone/s/data-availability.

“If there are ethical or legal restrictions on sharing a sensitive data set, authors should provide the following information within their Data Availability Statement upon submission:

• Explain the restrictions in detail (e.g., data contain potentially identifying or sensitive patient information)

• Provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent”

“Please note it is not acceptable for an author to be the sole named individual responsible for ensuring data access.”

If your university has imposed restrictions on sharing of this data, you must provide the contact information for a data access committee (or equivalent institutional

committee) that is authorized to review data requests and share dataupon approval. One of the authors cannot be named as being responsible for data requests.

Re: According to this comment, we have added the relevant information as supporting data in files S1 to S4.

Reviewer: 1

All the comments and concerns have been satisfactorily addressed except for the data availability issue.

Re: Thank you for appreciating our efforts. We have added the requested information as supporting data in files S1 to S4.

Reviewer: 2

1. Data Sharing (Major concern): The authors have not followed the PLOS Data policy which requires authors to make all data underlying the findings described in their manuscript fully available without restriction. The authors cited “potentially sensitive patient information” as the rationale for not making the data available. It is unclear how deidentified patients’ data contain any potentially sensitive patient information. It is also questionable whether a data access request from a fellow researcher will be honored in the future.

Re: Initially, we thought that the data could not be made public because they were clinical data. However, as you point out, the data are completely anonymized and thus not individuals are not identifiable. Therefore, we have made the data available by including it as supporting information in files S1 to S4.

Introduction

2. The added hypothesis was “dysbiosis of the microbiota is associated with pathogenesis of PMDs.” Please define “dysbiosis.” Please stay away from causal language as this small cross-sectional study could not establish causation.

Re: As stated in the limitations, we understand that it is not possible to state causality based on the results of this study. We have revised the manuscript to avoid any causal language. The purpose of our study was to explore the relationship between the gut microbiota and PMDs in an exploratory manner. To make this clear, we have revised the text on page 5, lines 87–89.

3. The authors used the term “low grade inflammation.” What does it mean? What are the characteristics/markers of low grade inflammation?

Re: In reference 22, low grade inflammation is defined as a CRP level of greater than 3 mg/L. We have added this information on page 5, lines 75–76.

Methods:

4. I am puzzled by the fact that 7 symptom-free women reported “symptom interference with social life”. Please

explain.

Re: We apologize for the lack of clarity in our explanation. As we describe in the Introduction on page 4, line 56, the prevalence of premenstrual symptoms is high (80%–90%), so completely symptom-free women are rare. The inclusion criteria for the N group included no symptoms listed in the PSQ showing a moderate or higher level, as described on page 6, lines 112–113, so this means that the N group includes those with mild premenstrual symptoms. These women are not completely free of premenstrual symptoms. The seven patients excluded from the N group had multiple mild premenstrual symptoms and mild interference of social life due to these symptoms as assessed by the PSQ. To clarify, we have revised the text on page 7, lines 124–128. In addition, we changed the wording of the description of the N group in Fig. 1 from “Not reporting premenstrual symptoms” to “Not reporting serious premenstrual symptoms”. For consistency, the text on page 2, line 26, 28 and on page 6, line 105, was also changed accordingly.

5. Some details on sample collection are missing. What is the length of time from collection to receipt by the lab? Participant compliance can be an issue when you ask participants to store stool samples in 4 Celsius degrees (in their fridge??). There can be issues related to sample integrity. Please comment how you approached participant compliance or acknowledge this as a limitation.

Re: In this study, special sampling tubes were used for DNA collection from stool specimens. Within these tubes was a storage solution containing guanidine, which is guaranteed to be stable for one month at room temperature or 4 °C; therefore, subjects did not need to store their specimens in a refrigerator. We have added this information on page 9, lines 177–182.

6. Please describe the rationale to assess bacteria translocation indirectly rather than directly.

Re: Analyzing the DNA of intestinal bacteria in blood is more labor-intensive and costly than indirect marker measurements. Given the exploratory nature of this study, we chose to examine bacterial translocation via an indirect method. We have added this explanation on page 18, lines 342–344.

7. Those who took drospirenone-containing oral contraceptives were excluded from the study. What about other types of hormonal contraceptives? Did you collect the data on those and consider them when interpreting the findings?

Re: Drospirenone-containing OCs were excluded because, unlike other conventional OCs, they have a proven therapeutic effect on PMDs. The percentage of Japanese women who use OCs as a contraceptive method is low (0.9%, as reported by the WHO in 2015), and as a result, none of the subjects in this study were OC users. We have added additional information to explain this on page 6, lines 110–111 and 117.

Results

8. It is unclear why LEfSE plots could not be used.

Re: In the first revision, we seem to have misunderstood the reviewer's comments on “LEfSE plots”. Assuming that "LEfSE plots" refers to cladograms, we have added these data as a new Fig 4A. We also changed the figure legend for Fig 4 accordingly on page 15, lines 300–302 and page 16, lines 303–306.

.

9. Can you report the effect size between groups for gut microbiome data?

Re: We analyzed the effect sizes and have added these data to Fig 3.

Discussion

10. Please explain the negative findings regarding alpha diversity.

Re: Alpha diversity was already described in the original version of the study in the case of MDD and inflammatory bowel disease, but a comparison with the PMD results obtained here was missing. Therefore, an explanation of alpha diversity relating to PMDs and MDD was added on page 18, line 354.

Overall

11. The manuscript can benefit from proof-reading by an English

editor.

Re: Our revised manuscript has been reviewed by a native English-speaking expert editor.

We hope our revised manuscript is now acceptable for publication in PLOS ONE. We look forward to hearing from you at your earliest convenience.

Yours sincerely,

Takashi Takeda, MD, PhD

Submitted filename: Response to Reviewrs.docx

Click here for additional data file.

4 Apr 2022

28 Apr 2022

Professor Emily Chenette

Editor-in-Chief

PLOS ONE

April 28, 2022

Dear Professor Chenette,

Thank you for reconsidering our manuscript entitled “Characteristics of the gut microbiota in women with premenstrual symptoms: A cross-sectional study” as an Original Research article for publication in PLOS ONE. We have revised our manuscript in response to the helpful comments of the editor, as shown below.

Additional Editor Comments:

The authors have adequately addressed the comments on the prior revision of their manuscript, with the exception that the data made available consists of derived datasets (genus and phylum level count table) rather than raw sequence data as is standard in the field (i.e. fastq files). Could the authors deposit the raw sequence data in a public repository (e.g. NCBI SRA) or provide a reason why the raw sequence data are no longer available?

Re: Following this comment, we registered the raw sequence data to the public repository (DRA) and added the accession numbers on page 23, line 442–445.

In this revision, we reviewed the description of the method section and found some mistakes. Therefore, we have revised it as described on page 10, line 202–204.

We hope our revised manuscript is now acceptable for publication in PLOS ONE. We look forward to hearing from you at your earliest convenience.

Yours sincerely,

Takashi Takeda, MD, PhD

Submitted filename: Response to Reviewrs R3.docx

Click here for additional data file.

2 May 2022

Characteristic of the gut microbiota in women with premenstrual symptoms : a cross-sectional study

PONE-D-21-30216R3

Dear Dr. Takeda,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Jonathan Jacobs

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

17 May 2022

PONE-D-21-30216R3

Characteristics of the gut microbiota in women with premenstrual symptoms: a cross-sectional study

Dear Dr. Takeda:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Jonathan Jacobs

Academic Editor

PLOS ONE