Genome-Wide Association Study of Distressing Premenstrual Symptoms in Two Nordic Populations

58

59 Premenstrual disorders (PMDs) are characterized by affective and physical 60 symptoms before menses, likely due to abnormal sensitivity to normal hormone 61 fluctuations. While sizable heritability has been indicated in twin studies, there are no 62 genome-wide association studies (GWAS) to inform the genetic architecture of 63 PMDs. 64 65

66 We conducted a GWAS of 17,511 women with distressing premenstrual symptoms 67 (DPS) and 54,789 women controls of European ancestry from two Nordic population-68 based cohorts. DPS were assessed using questionnaire or identified as a clinical 69 diagnosis of PMDs in the nationwide healthcare registers. GWAS was performed in 70 each study before meta-analysis, analyses of single nucleotide polymorphism (SNP)-71 based heritability (h2 SNP) and genetic correlations to psychosocial and gynecological 72 phenotypes, as well as blood levels of gonadal steroids. 73 74

75 In the meta-analysis, one locus at 12p13.3 (rs758170, CACNA1C, P=1.53x10-8, 76 OR=0.93, 95% CI 0.90-0.95) was associated with DPS. The SNP-based heritability 77 was estimated 0.072 (SE=0.01, P=2.46 x10-12). Statistically significant genetic 78 correlations (rg) were found between DPS and all major psychiatric disorders, with 79 the strongest correlation with major depression (rg=0.62, CI 0.49-0.74, P=3.04x10-80 22). Weaker correlations were noted to gynecological conditions such as 81 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 5 endometriosis (rg=0.17, CI 0.01-0.32, P=0.029), while gonadal steroid hormone 82 levels in blood were uncorrelated. 83 84

85 This study provides the first direct insights into the genetic architecture of PMDs by 86 identifying a SNP associated with DPS and genetic correlations to other conditions. If 87 confirmed in larger independent populations, these findings may advance our 88 understanding of the underlying mechanisms of PMDs. 89 90

GWAS, premenstrual disorders, genetic correlation, heritability, women’s 91 health, MoBa, LifeGene 92 93 94 95 96 97 98 99 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 6

100 Premenstrual disorders (PMDs), encompassing premenstrual syndrome (PMS) and 101 the more severe premenstrual dysphoric disorder (PMDD), are characterized by 102 substantial mood, behavioral, and physical symptoms that occur in the luteal phase 103 of the menstrual cycle and resolve with the onset of menstruation1. PMDs affect 104 millions of women of reproductive age worldwide, with an estimated prevalence of 105 20%-30% for PMS and 2%-6% for PMDD1. Both severe PMS and PMDD are 106 accompanied by significantly impaired social activities and relationships2–4. Although 107 the symptoms are often limited to the days before menses, the chronic and cyclical 108 nature of PMDs can have a significant impact on a woman’s life5, including 109 increasing the risk of suicidal behavior6,7. 110 111 While epidemiological studies have highlighted potential links with other hormone-112 related conditions (e.g., perinatal depression (PND)8, menopause timing and 113 menopause symptoms9), an experimental study has revealed that PMDs are 114 triggered by an abnormal, or heightened, response to normal hormone fluctuations10. 115 However, the ontogeny of the abnormal response remains unknown, impeding the 116 development of new treatments. Thus, uncovering the underlying causes of PMDs is 117 essential for more effective detection and intervention. 118 119 Twin studies suggest a genetic component in PMDs, with an estimated heritability of 120 35%-57% for premenstrual symptoms11–15. Moreover, research into candidate genes 121 have indicated variants in estrogen receptor genes16 serotonin receptor 1A17, 122 transcription factor AP-2 beta18, and steroid-5-alpha-reductase,19 alpha polypeptide 123 119 in PMDD, although conflicting results also have been noted20. A genome-wide 124 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 7 approach opens for a comprehensive, hypothesis-free characterization of genetic 125 factors involved in PMDs, yet such an undertaking has not yet been performed for 126 PMDs. Here, we conducted a genome-wide association study (GWAS) of distressing 127 premenstrual symptoms (DPS) in two European-ancestry samples, aiming to identify 128 the genetic architecture and relationship of DPS to other psychiatric conditions and 129 underlying mechanisms of PMDs. 130 131

AND MATERIALS 132 Study Population 133 We conducted a GWAS of 72,297 European-ancestry women nested from the 134 LifeGene cohort in Sweden and Mother, Father and Child Cohort Study (MoBa) 135 cohort in Norway. 136 137 LifeGene is a large-scale Swedish prospective cohort launched in 2009 with 138 longitudinal follow-ups21. It enrolled 39,862 people (24,265 women) aged 18-50 139 years who were randomly selected from the Swedish population and their household 140 members. A thorough web-based questionnaire collecting information on lifestyle, 141 physical, mental, and social well-being was administered at baseline and in five 142 annual follow-up cycles. Blood samples were collected during the in-person testing 143 at baseline. Participants were linked to the national population and health registers 144 using their unique Swedish personal identification number, a lifelong identifier 145 assigned at birth or upon immigration to Sweden. Informed consent was obtained 146 either electronically from all participants upon registration online or in writing at the 147 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 8 in-person testing center. The current study was approved by the Swedish Ethical 148 Review Authority (2021-02775). 149 150 MoBa is a population-based pregnancy cohort study conducted by the Norwegian 151 Institute of Public Health22,23,24. Participants were recruited from across Norway from 152 1999 to 2008. Women consented to participation in 41% of pregnancies. The cohort 153 includes approximately 114,500 children, 95,200 mothers and 75,200 fathers. The 154 current study is based on version 12 of the quality-assured data files released for 155 research in January 2019. The establishment of MoBa and initial data collection 156 were approved via a license from the Norwegian Data Protection Agency and after 157 review of the Regional Committees for Medical and Health Research Ethics. The 158 MoBa cohort is currently regulated by the Norwegian Health Registry Act. The 159 current study was approved by The Regional Committees for Medical and Health 160 Research Ethics (2016/1226/REK). 161 162 Assessment of premenstrual symptoms 163 Both questionnaire assessments and nationwide healthcare registers were sourced 164 for case assessment. Cases with DPS were defined as either having a clinical 165 diagnosis of PMD recorded in the registers or having met the criteria for a probable 166 PMD based on self-report questionnaires (as described in detail below). The controls 167 were women with no clinical diagnosis of PMDs in registers and not meeting the 168 PMD criteria in all available questionnaire cycles. 169 170 LifeGene 171 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 9 A modified version of the Premenstrual Symptom Screening Tool (PSST)25 was 172 employed to assess premenstrual symptoms at baseline and annual follow-ups for 5 173 years26. The original PSST has been validated with a sensitivity of 79%27. The PSST 174 was modified to start with three screening questions: 1) “During most menstruation 175 cycles during the last year, have you experienced mood changes and/or physical 176 symptoms during the week before menstruation?”; 2) “Have your premenstrual 177 symptoms been so severe that they have affected your relationships with others or 178 your ability to perform work or other activities?”; and 3) “Are you absolutely certain 179 that the symptoms are limited to the premenstrual period, meaning that you are 180 always completely symptom-free approximately a week after menstruation begins?” 181 Upon confirmation of all screening questions, participants were prompted to rate the 182 severity of 15 physical and affective symptoms from 1 (none), 2 (moderate), 3 183 (considerably severe) to 4 (severe). As described elsewhere26, participants were 184 classified as cases if they met (1) ≥ 1 out of 4 affective symptoms rated as 185 considerably severe to severe; and (2) ≥ 4 other symptoms rated as moderate to 186 severe. 187 188 To complement the questionnaire assessment in LifeGene, for the Swedish 189 participants, we further identified clinical diagnosis of PMDs, as described 190 elsewhere7. According to the Swedish guidelines, a clinical diagnosis of PMDs 191 should be based on prospective daily symptom ratings for at least 2 consecutive 192 menstrual cycles28. Briefly, we identified PMD diagnoses based on ICD codes 193 (Supplementary Table S1) from the National Patient Register (1987-2023) and the 194 Stockholm Primary Care Register (2001-2021), as 82% of the participants lived in 195 Stockholm County. Since primary care data was unavailable for residents in other 196 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 10 counties, we also obtained information on filled prescriptions of antidepressants 197 (ATC codes: N06AB, N06AX, N06AA) and hormonal contraceptives (G03A) with a 198 written indication for PMD treatment from the National Prescribed Drug Register 199 (2006-2023). 200 201 MoBa 202 Questionnaire assessment was based on the women’s responses to the following 203 questions at 15 th week of gestation: “Are you usually depressed or irritable before 204 your period?” and “If yes, does this feeling disappear after you get your period?”. As 205 described elsewhere 29, the cases were defined based on the response “yes, 206 noticeably” or “yes, very much” to the first question, and “yes” to the second 207 question. We excluded individuals whose symptoms of PMD did not resolve after the 208 onset of menses. 209 210 Clinical diagnoses of PMDs were derived from the Primary Care Registry of Norway 211 (KUHR; years 2006-2023), which contains diagnoses given at the primary care level. 212 We identified cases based on the Premenstrual Tension Syndrome (X89) diagnosis 213 according to the International Classification of Primary Care, Second Edition (ICPC-214 2)30. 215 216 Genotyping, quality control, and imputation 217 LifeGene 218 In LifeGene, DNA was extracted from blood samples collected at baseline. 219 Participants from the LifeGene were genotyped by four sub-studies including the 220 present study (Supplementary Table S2). Quality control (QC) was performed using 221 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 11 the Ricopili bioinformatics pipeline for each sub-study31. Briefly, these steps included 222 a call rate threshold of ≥ 0.98 for cases and controls, heterozygosity (FHET) within 223 ±0.20, and exclusion of sex mismatches. SNP QC required a call rate of ≥ 0.98, 224 missingness difference ≤ 0.02 and minor allele frequency (MAF) ≥ 0.01. Ungenotyped 225 SNPs were then imputed based on the Haplotype Reference Consortium (HRC) 226

panel (r1.1)32 via the Sanger imputation server and pooled after removing 227 duplicated individuals. Information on a total of 7,135,674 single-nucleotide 228 polymorphisms (SNPs) was thereby available for analysis. Admixture analysis was 229 performed using the ADMIXTURE software to estimate genetic ancestry proportions, 230 leveraging reference populations from the 1000 Genomes Project (phase 3 v5)33. 231 Analysis was restricted to autosomal variants. Individuals estimated with less than 232 90% probability of European ancestry were excluded. Briefly, genotype data were 233 available for 8,826 women (36%); after excluding 862 participants who were first-234 degree relatives (having identity-by-descent sharing ≥ 0.2), had significant non-235 European ancestry (n=896), or had no information on phenotype (n=1,839), 5,229 236 participants were included in this analysis. 237 238 In MoBa, venous blood was collected from women around the 15th week of 239 gestation and immediately after giving birth. Genomic DNA is extracted and stored at 240 the Norwegian Institute of Public Health 34. The MoBa cohort genotyping was 241 conducted through multiple research projects over several years35. A novel family-242 based pipeline (MoBaPsychGen genotype QC pipeline) was implemented to handle 243 the relatedness structure of the MoBa dataset, while appropriately accounting for the 244 differences resulting from array and batch effects35. The pipeline36 includes pre-245 imputation QC, phasing, imputation, and post-imputation QC, and prioritizes 246 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 12 retaining individuals over SNPs. After QC procedure 6,981,748 SNPs were available 247 for further analysis. Our analysis was restricted to individuals of European ancestry, 248 selected based on visual comparison of the first seven genetic principal components 249 (PCs) with PCs from 1000 Genomes phase 1 unrelated samples. 250 251 252 Statistical analysis 253 GWAS 254 Characteristics of individuals with and without DPS (age, educational level, civil 255 status, depression, anxiety disorder, age at menarche and parity) in LifeGene and 256 MoBa were summarized using descriptive statistics. P-values were calculated using 257 chi-square test. 258 259 In LifeGene, GWAS was performed using logistic regression using PLINK237. 260 Analysis was restricted to 6,508,434 genetic variants with a minor allele frequency 261 (MAF) ≥ 0.01 and imputation quality score (INFO) ≥ 0.90. In MoBa, logistic 262 regression was employed for GWAS analysis using Regenie (version 3.2.5)38. We 263 excluded SNPs with imputation info score (minINFO) < 0.80 and SNPs with minor 264 allele count (MAC) below 20. For both cohorts, the estimates were adjusted for sub-265 study membership and the top 10 PCs in Model 1. To reduce the genetic influence 266 of other common psychiatric conditions, model 2 was additionally adjusted for clinical 267 diagnosis of depression and anxiety disorder (ICD codes in Supplementary Table 268 S1). In an additional analysis, we restricted analysis to cases confirmed by both 269 questionnaire assessment and clinical diagnosis. 270 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 13 We generated the Q-Q plot using qqman39 package in R version 4.2.2 (2022-10-31). 271 Summary statistics from both LifeGene and MoBa were then meta-analyzed using 272 inverse variance-weighted analysis as implemented in METAL (version 2020-05-273 05)40. SNPs with P-value < 5x10-8 were considered genome-wide significant and P-274 value < 1x10-6 for borderline significance. The odds ratio (OR), 95%CI, and P-value 275 were reported for any independent SNPs (LD r2<0.6) above marginal significance 276 (i.e., lead SNPs). Functional annotation was conducted using FUMA41 integrating 277 eQTL data from brain and blood tissues. 278 279 SNP-based heritability and genetic correlation 280 Based on the summary statistics, we estimated the SNP-based heritability on the 281 observed scale of DPS using the linkage disequilibrium score regression (LDSC) 282 software package (version 2.0.1)42. Several studies have suggested that PMDs are 283 linked to a broad range of other disorders and traits43–45. We used LDSC to 284 undertake genetic correlation analyses with major psychiatric traits (e.g., major 285 depression), gynecologic conditions (e.g., endometriosis), sex hormones (e.g., 286 absolute level of estradiol in the blood), and known risk factors (e.g., age at 287 menarche) for PMDs as described in the Supplementary Table S3. 288 289

290 A total of 72,297 (5,229 from LifeGene and 67,068 from MoBa) women were 291 included in the final GWAS meta-analysis. Among them, 17,511 (24.2%) met the 292 criteria for DPS (1,962 (37.5%); cases were oversampled for genotyping, while the 293 prevalence in the whole cohort was 10.8%) from LifeGene and 15,549 (23.1%) from 294 MoBa. In LifeGene, compared to the controls, women with DPS were more likely to 295 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 14 have lower educational attainment, not to have a partner, to have prior history of 296 depression or anxiety disorder (Table 1). Similar characteristics were observed 297 among MoBa participants. 298 299 GWAS analysis 300 Quantile-quantile analysis of the GWAS meta-analysis indicated moderate inflation 301 (λ =1.075, Supplementary Tables Figure S1). One locus at 12p13.3 was genome-302 wide significant (rs758170, P=1.53x10-8, OR=0.93, 95% CI 0.90–0.95, risk allele C) 303 (Figure 2 and Table 2). The top SNP rs758170 (intron) was positionally mapped to 304 CACNA1C (Calcium Voltage-Gated Channel Subunit Alpha1 C) gene. FUMA, 305 indicates that the C allele associated with DPS risk leads to increased expression of 306 CACNA1C in cerebellum (P=2.53x10-6) (Figure 3). Comparable point estimates for 307 rs758170 were found between LifeGene (OR = 0.95) and MoBA (OR = 0.92, P-for-308 heterogeneity = 0.59). Moreover, six other loci showed a suggestive level of 309 significance (P<1.0x10-6, Table 2). 310 311 Next, we conducted several sensitivity analyses using LifeGene samples to test the 312 robustness of our findings. To reduce the influence of comorbid psychiatric disorders, 313 we additionally adjusted for depression and anxiety, yielding comparable 314 associations for top loci (Supplementary Table S5). To minimize the misclassification 315 of probable PMD cases, we restricted analysis to cases confirmed by both 316 questionnaire assessment and clinical diagnosis (n= 762) and found largely 317 comparable point estimates despite the attenuation of P values due to smaller 318 sample size (Supplementary Table S6). 319 320 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 15 SNP heritability and genetic correlation 321 The SNP-based heritability was estimated as 0.072 (SE=0.01, P=2.46x10-12). We 322 observed significant positive genetic correlations between DPS and all considered 323 psychiatric disorders (rg= 0.32 - 0.62), with the strongest correlation with major 324 depression (rg= 0.62, 95%CI = (0.49, 0.74), empirical P= 3.04x10-22, Figure 2). There 325 was a significant positive genetic correlation between PMDs and endometriosis (rg= 326 0.17, 95%CI = (0.01 - 0.32), empirical P= 0.029). In addition, a positive genetic 327 correlation was found for BMI (rg= 0.1, 95%CI = (0.03, 0.17), empirical P= 0.003), 328 while significant negative genetic correlations were observed with age at first 329 childbirth (rg= -0.35, 95%CI = (-0.47, -0.22), empirical P= 7.01x10-8), subjective well-330 being (rg= -0.34, 95%CI = (-0.50, -0.19), empirical P= 1.82x10-5), and educational 331 attainment (rg= -0.15, 95%CI = (-0.22, -0.07), empirical P= 8.39x10-5) 332 (Supplementary Table S4). Finally, we did not observe any significant correlation with 333 blood steroid hormone levels. 334 335

336 To our knowledge, this is the first GWAS focused on DPS, including 17,511 337 individuals with DPS and 54,789 control individuals of European genetic ancestry. 338 We identified a genome-wide significant risk locus on 12p13.3, with comparable 339 point estimates in LifeGene and MoBa. A moderate SNP-based heritability of 7.2% 340 was observed. Genetic correlations were found between DPS and a range of 341 psychiatric disorders, with the strongest genetic overlap noted for major depression. 342 343 The CACNA1C gene implicated by linkage to psychiatric disorders, encodes the 344 calcium voltage-gated channel subunit alpha1C, crucial for calcium channel 345 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 16 functioning essential to neurodevelopment46,47. Previous research has found that the 346 gene NUCB1, encoding a calcium-binding protein, is directly involved in regulating 347 intracellular Ca2+ within the endoplasmic reticulum (ER)-Golgi compartment, 348 contributing to the abnormal response to steroid hormones in patients with PMDD48. 349 Moreover, studies on calcium signalling in neural cells suggest potential interactions 350 between KCNMA1 (calcium-dependent gene) and NUCB1 and CACNA1C, as part of 351 calcium regulation pathways49,50. Since Ca2+ activity in developing neural cells is 352 modulated by various membrane receptors, including GABA, these findings highlight 353 promising links between calcium regulation and GABA receptor function, which has 354 been widely studied in relation to PMDs51,52. 355 356 CACNA1C has been linked to a range of psychiatric disorders53,54. In a study 357 investigating the effects of CACNA1C haploinsufficiency on mouse behavior in tests 358 with relevance to human mood disorders, researchers found that an intronic region 359 of CACNA1C was involved in mood disorder pathophysiology55. Moreover, several 360 association studies have linked polymorphisms in CACNA1C to bipolar disorder and 361 schizophrenia53,54, potentially through variations in mean gray matter volume and 362 medio temporal emotional processing56,57. Indeed, rs758170 is highly correlated (LD 363 r2>0.8) with variants that have been associated with bipolar disorder58,59, 364 schizophrenia60, autism and attention deficit hyperactivity disorder (ADHD)61, 365 suggesting potentially shared disease mechanisms between PMDs and these 366 psychiatric disorders. However, the association of rs758170 remains similar after 367 adjustment for depression and anxiety, indicating that our finding cannot be 368 completely explained by psychiatric comorbidities. In addition, rs758170 has been 369 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 17 linked to lipid metabolism62, for which several epidemiological studies have illustrated 370 a salient relationship between adiposity and PMDs63,64. 371 372 Previous twin research indicates a sizable genetic contribution to premenstrual 373 symptoms15, our study represents the first attempt to estimate the genetic liability of 374 probable PMDs. As reported in previous genetic studies of other complex 375 traits65,66,67, the SNP-based heritability estimate in our study (h2 SNP = 0.07) is lower 376 than reported in twin studies. Future research should focus on larger GWAS as well 377 as rare alleles, other alleles not well captured by GWAS and gene-environment 378 interplay to better capture the genetic liability of PMDs. 379 380 Extensive clinical and questionnaire-based research consistently demonstrate a high 381 prevalence of psychiatric comorbidities, particularly depressive and anxiety 382 disorders, among individuals diagnosed with premenstrual disorders1,68. Additionally, 383 other conditions such as bipolar disorder and ADHD have been found to co-occur 384 with PMDs44,69. In the present analyses, we found the largest and most significant 385 genetic correlation of PMDs with major depression, aligning with the confirmed 386 phenotypic correlation in literature70. In addition, we observed genetic correlations 387 with ADHD and bipolar disorder (including bipolar I, bipolar II, and schizoaffective 388 type), echoing the previous report on the polygenetic association with these 389 disorders observed in MoBa29. Moreover, we report a significant genetic correlation 390 between endometriosis and DPS. This supports previous research conducted on 391 endometriosis and the phenotypic relationship with major depression and other 392 psychiatric disorders, potentially explained by shared dysregulated immunological 393 functions71. Finally, no genetic correlation was observed with steroid hormone levels, 394 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 18 lending support to the notion that PMDs are not driven by hormone levels but 395 sensitivity to these fluctuations as shown in previous experimental studies72. 396 397 This study is strengthened by its large sample size, the inclusion of two Nordic 398 cohorts with banked bio samples, and the use of rich register and questionnaire data 399 for case identification. This study also has some limitations. First, a portion of the 400 individuals with DPS were identified through screening tools. However, prospective 401 symptom charting, as required to establish the diagnosis, is not feasible in cohort 402 studies and may result in a high attrition rate particularly for those with severe 403 symptoms73. Moreover, to complement on questionnaire assessment, we have used 404 clinical diagnoses recorded from national and regional healthcare registers. While we 405 lacked information on clinical diagnostic process, prospective symptom charting has 406 been outlined in clinical guidelines in Sweden28. With both questionnaire assessment 407 and register-recorded diagnoses, we may have captured moderate/severe PMS, 408 PMDD, and false positive. In a sensitivity analysis, we however restricted cases to 409 those confirmed by both questionnaire assessment and clinical diagnosis, 410 presumably with a higher validity, and yielded comparable point estimates for top 411 SNPs. Secondly, due to small numbers, we removed the participants of non-412 European ancestry. Future studies should include diverse ancestral backgrounds to 413 aim for generalizable results across different groups, making the findings relevant 414 and potentially beneficial to a wider population, as the lack of replication in the 415 current study is a limitation. 416 417 In summary, in this first genome-wide association meta-analysis of DPS, we report a 418 significant locus and genetic correlations with a range of psychosocial and 419 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 19 gynecological phenotypes. The identified genetic markers may help advance our 420 understanding of the underlying mechanisms, which may further aid early detection 421 and clinical management of PMDs. Future studies with larger, more diverse samples, 422 and cases confirmed with prospective symptom charting are needed to further 423 understand the genetic influence on PMDs, and potentially the mechanisms of mood 424 regulating effects of sex hormones. 425 426 427

428 This work was funded by the Swedish Research Council (grant number 2020-01003 429 and 2024-02592 to DL), European Research Council (101165552 to DL), Swedish 430 Research Council for Health, Working Life and Welfare (2023-00399 to DL), 431 Karolinska Institutet (to DL), and Karolinska Institutet - National Institute of Health 432 Neuroscience Doctoral Program (to EH, DL, and PJS). OAA received support from 433 the Research Council of Norway (324252, 324499, 300309, 326813, 334920, 434 271555/F21), the South-East Norway Regional Health Authority (2023-031 and 435 2022-073), The University of Oslo, KG Jebsen Stiftelsen (SKGJ-MED-021), the 436 European Union’s Horizon 2020 research and innovation program (847776, 964874), 437 NordForsk (164218) and the US National Institutes of Health (R01MH123724-01). 438 CMB was supported by Swedish Research Council (Vetenskapsrådet, award: 538-439 2013-8864). YL was supported by US NIMH R01 MH123724, the Swedish Research 440 Council (2021-02615), and the European Research Council (101042183). ADF was 441 supported by European Research Council (947763). We extend our gratitude to the 442 LifeGene and MoBa participants for their time and commitment, which made this 443 research possible. 444 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 20 445 LifeGene was supported by grants from the Swedish Research Council, Karolinska 446 Institutet, Karolinska Institutet/Stockholm County Council core facility funds, the 447 Ragnar and Torsten Söderberg Foundation, and AFA Insurance. Genotyping was 448 performed by the SNP&SEQ Technology Platform in Uppsala (www.genotyping.se). 449 The facility is part of the National Genomics Infrastructure supported by the Swedish 450 Research Council for Infrastructures and Science for Life Laboratory, Sweden. The 451 computations/data handling/[SIMILAR] were/was enabled by resources provided by 452 the National Academic Infrastructure for Supercomputing in Sweden (NAISS), 453 partially funded by the Swedish Research Council through grant agreement no. 454 2022-06725. 455 456 MoBa is supported by the Norwegian Ministry of Health and Care Services and the 457 Ministry of Education and Research. We are grateful to all the participating families 458 in Norway who take part in this on-going cohort study. We thank the Norwegian 459 Institute of Public Health (NIPH) for generating high-quality genomic data. This 460 research is part of the HARVEST collaboration, supported by the Research Council 461 of Norway (grant 229624). The Norwegian Centre for Mental Disorders Research 462 (NORMENT) provided genotype data, funded by the Research Council of Norway 463 (grant 223273), South East Norway Health Authorities and Stiftelsen Kristian 464 Gerhard Jebsen. We further thank the Center for Diabetes Research, the University 465 of Bergen for providing genotype data and performing quality control and imputation 466 of the data funded by the ERC AdG project SELECTionPREDISPOSED, Stiftelsen 467 Kristian Gerhard Jebsen, Trond Mohn Foundation, the Research Council of Norway, 468 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 21 the Novo Nordisk Foundation, the University of Bergen, and the Western Norway 469 Health Authorities. 470 471 We would like to thank Robert Karlsson, PhD (https://orcid.org/0000-0002-8949-472 2587) for his invaluable assistance with Ricopili and data management, which were 473 crucial to the success of this study. 474 475 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 22

476 1. Yonkers KA, Simoni MK. Premenstrual disorders. American Journal of Obstetrics 477 and Gynecology. 2018;218(1):68-74. doi:10.1016/j.ajog.2017.05.045 478 2. O’Brien PMS, Bäckström T, Brown C, et al. Towards a consensus on diagnostic 479 criteria, measurement and trial design of the premenstrual disorders: the ISPMD 480 Montreal consensus. Arch Womens Ment Health. 2011;14(1):13-21. 481 doi:10.1007/s00737-010-0201-3 482 3. Guidelines for Women’s Health Care: A Resource Manual. American College of 483 Obstetricians and Gynecologists, Women’s Health Care Physicians; 2014. 484 4. Diagnostic and Statistical Manual of Mental Disorders | Psychiatry Online. DSM 485 Library. Accessed October 3, 2024. 486 https://psychiatryonline.org/doi/book/10.1176/appi.books.9780890425596 487 5. Wei MY , Kawachi I, Okereke OI, Mukamal KJ. Diverse Cumulative Impact of 488 Chronic Diseases on Physical Health-Related Quality of Life: Implications for a 489 Measure of Multimorbidity. Am J Epidemiol. 2016;184(5):357-365. 490 doi:10.1093/aje/kwv456 491 6. Yang Q, Sjölander A, Li Y, et al. Clinical indications of premenstrual disorders and 492 subsequent risk of injury: a population-based cohort study in Sweden. BMC 493 Medicine. 2021;19(1):119. doi:10.1186/s12916-021-01989-4 494 7. Opatowski M, Valdimarsdóttir UA, Oberg AS, Bertone-Johnson ER, Lu D. 495 Mortality Risk Among Women With Premenstrual Disorders in Sweden. JAMA 496 Network Open. 2024;7(5):e2413394. doi:10.1001/jamanetworkopen.2024.13394 497 8. Yang Q, Bränn E, Johnson ERB, et al. The bidirectional association between 498 premenstrual disorders and perinatal depression: A nationwide register-based 499 study from Sweden. PLOS Medicine. 2024;21(3):e1004363. 500 doi:10.1371/journal.pmed.1004363 501 9. Yang Y , Valdimarsdóttir UA, Manson JE, et al. Premenstrual Disorders, Timing of 502 Menopause, and Severity of Vasomotor Symptoms. JAMA Network Open. 503 2023;6(9):e2334545. doi:10.1001/jamanetworkopen.2023.34545 504 10. Schmidt PJ, Nieman LK, Danaceau MA, Adams LF, Rubinow DR. Differential 505 behavioral effects of gonadal steroids in women with and in those without 506 premenstrual syndrome. N Engl J Med. 1998;338(4):209-216. 507 doi:10.1056/NEJM199801223380401 508 11. Kendler KS, Silberg JL, Neale MC, Kessler RC, Heath AC, Eaves LJ. Genetic 509 and environmental factors in the aetiology of menstrual, premenstrual and 510 neurotic symptoms: a population-based twin study. Psychol Med. 1992;22(1):85-511 100. doi:10.1017/s0033291700032761 512 12. Kendler KS, Karkowski LM, Corey LA, Neale MC. Longitudinal population-based 513 twin study of retrospectively reported premenstrual symptoms and lifetime major 514 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 23 depression. Am J Psychiatry. 1998;155(9):1234-1240. 515 doi:10.1176/ajp.155.9.1234 516 13. van den Akker OB, Eves FF, Stein GS, Murray RM. Genetic and environmental 517 factors in premenstrual symptom reporting and its relationship to depression and 518 a general neuroticism trait. J Psychosom Res. 1995;39(4):477-487. 519 doi:10.1016/0022-3999(94)00152-u 520 14. Treloar SA, Heath AC, Martin NG. Genetic and environmental influences on 521 premenstrual symptoms in an Australian twin sample. Psychol Med. 522 2002;32(1):25-38. doi:10.1017/s0033291701004901 523 15. Jahanfar S, Lye MS, Krishnarajah IS. The Heritability of Premenstrual Syndrome. 524 Twin Research and Human Genetics. 2011;14(5):433-436. 525 doi:10.1375/twin.14.5.433 526 16. Huo L, Straub RE, Roca C, et al. Risk for premenstrual dysphoric disorder is 527 associated with genetic variation in ESR1, the estrogen receptor alpha gene. Biol 528 Psychiatry. 2007;62(8):925-933. doi:10.1016/j.biopsych.2006.12.019 529 17. Dhingra V, Magnay JL, O’Brien PMS, Chapman G, Fryer AA, Ismail KMK. 530 Serotonin receptor 1A C(-1019)G polymorphism associated with premenstrual 531 dysphoric disorder. Obstet Gynecol. 2007;110(4):788-792. 532 doi:10.1097/01.AOG.0000284448.73490.ac 533 18. Damberg M, Westberg L, Berggård C, et al. Investigation of transcription factor 534 AP-2 beta genotype in women with premenstrual dysphoric disorder. Neurosci 535 Lett. 2005;377(1):49-52. doi:10.1016/j.neulet.2004.11.068 536 19. Adams M, McCrone S. SRD5A1 genotype frequency differences in women with 537 mild versus severe premenstrual symptoms. Issues Ment Health Nurs. 538 2012;33(2):101-108. doi:10.3109/01612840.2011.625514 539 20. Melke J, Westberg L, Landén M, et al. Serotonin transporter gene polymorphisms 540 and platelet [3H] paroxetine binding in premenstrual dysphoria. 541 Psychoneuroendocrinology. 2003;28(3):446-458. doi:10.1016/s0306-542 4530(02)00033-1 543 21. Almqvist C, Adami HO, Franks PW, et al. LifeGene--a large prospective 544 population-based study of global relevance. Eur J Epidemiol. 2011;26(1):67-77. 545 doi:10.1007/s10654-010-9521-x 546 22. Magnus P , Irgens LM, Haug K, et al. Cohort profile: the Norwegian Mother and 547 Child Cohort Study (MoBa). Int J Epidemiol. 2006;35(5):1146-1150. 548 doi:10.1093/ije/dyl170 549 23. Magnus P , Birke C, Vejrup K, et al. Cohort Profile Update: The Norwegian Mother 550 and Child Cohort Study (MoBa). Int J Epidemiol. 2016;45(2):382-388. 551 doi:10.1093/ije/dyw029 552 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 24 24. Paltiel L, Anita H, Skjerden T , et al. The biobank of the Norwegian Mother and 553 Child Cohort Study – present status. Norsk Epidemiologi. 2014;24(1-2). 554 doi:10.5324/nje.v24i1-2.1755 555 25. Steiner M, Macdougall M, Brown E. The premenstrual symptoms screening tool 556 (PSST) for clinicians. Arch Womens Ment Health. 2003;6(3):203-209. 557 doi:10.1007/s00737-003-0018-4 558 26. Westermark V, Yang Y , Bertone-Johnson E, et al. Association between severe 559 premenstrual disorders and change of romantic relationship: A prospective cohort 560 of 15,606 women in Sweden. Journal of Affective Disorders. 2024;364:132-138. 561 doi:10.1016/j.jad.2024.08.032 562 27. Henz A, Ferreira CF, Oderich CL, et al. Premenstrual Syndrome Diagnosis: A 563 Comparative Study between the Daily Record of Severity of Problems (DRSP) 564 and the Premenstrual Symptoms Screening Tool (PSST). Revista Brasileira de 565 Ginecologia e Obstetrícia / RBGO Gynecology and Obstetrics. 2017;40:20-25. 566 doi:10.1055/s-0037-1608672 567 28. Riktlinjer vid premenstruell dysforisk störning, PMDS. March 13, 2023. Accessed 568 September 25, 2024. 569 https://janusinfo.se/behandling/expertgruppsutlatanden/kvinnosjukdomarochforlo570 ssning/kvinnosjukdomarochforlossning/riktlinjervidpremenstruelldysforiskstorning571 pmds.5.6081a39c160e9b387319f3.html 572 29. Jaholkowski P , Shadrin AA, Jangmo A, et al. Associations Between Symptoms of 573 Premenstrual Disorders and Polygenic Liability for Major Psychiatric Disorders. 574 JAMA Psychiatry. 2023;80(7):738-742. doi:10.1001/jamapsychiatry.2023.1137 575 30. World Organization of National Colleges, Academies, and Academic Associations 576 of General Practitioners/Family Physicians, ed. ICPC-2-R: International 577 Classification of Primary Care. Rev. 2nd ed. Oxford University Press; 2005. 578 31. Lam M, Awasthi S, Watson HJ, et al. RICOPILI: Rapid Imputation for COnsortias 579 PIpeLIne. Bioinformatics. 2020;36(3):930-933. doi:10.1093/bioinformatics/btz633 580 32. Loh PR, Danecek P , Palamara PF, et al. Reference-based phasing using the 581 Haplotype Reference Consortium panel. Nat Genet. 2016;48(11):1443-1448. 582 doi:10.1038/ng.3679 583 33. Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry 584 in unrelated individuals. Genome Res. 2009;19(9):1655-1664. 585 doi:10.1101/gr.094052.109 586 34. Protocols for the Norwegian Mother, Father and Child Cohort Study (MoBa). 587 Norwegian Institute of Public Health. May 3, 2019. Accessed September 25, 588 2024. https://www.fhi.no/en/publ/2012/protocols-for-moba/ 589 35. Corfield EC, Frei O, Shadrin AA, et al. The Norwegian Mother, Father, and Child 590 cohort study (MoBa) genotyping data resource: MoBaPsychGen pipeline v.1. 591 Published online June 26, 2022:2022.06.23.496289. 592 doi:10.1101/2022.06.23.496289 593 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 25 36. psychgen/MoBaPsychGen-QC-pipeline. Published online April 17, 2023. 594 Accessed September 25, 2024. https://github.com/psychgen/MoBaPsychGen-595 QC-pipeline 596 37. Purcell S, Neale B, Todd-Brown K, et al. PLINK: A Tool Set for Whole-Genome 597 Association and Population-Based Linkage Analyses. Am J Hum Genet. 598 2007;81(3):559-575. 599 38. Mbatchou J, Barnard L, Backman J, et al. Computationally efficient whole-600 genome regression for quantitative and binary traits. Nat Genet. 601 2021;53(7):1097-1103. doi:10.1038/s41588-021-00870-7 602 39. Turner SD. qqman: an R package for visualizing GWAS results using Q-Q and 603 manhattan plots. Journal of Open Source Software. 2018;3(25):731. 604 doi:10.21105/joss.00731 605 40. Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of 606 genomewide association scans. Bioinformatics. 2010;26(17):2190-2191. 607 doi:10.1093/bioinformatics/btq340 608 41. Watanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping 609 and annotation of genetic associations with FUMA. Nat Commun. 610 2017;8(1):1826. doi:10.1038/s41467-017-01261-5 611 42. Bulik-Sullivan BK, Loh PR, Finucane HK, et al. LD Score regression distinguishes 612 confounding from polygenicity in genome-wide association studies. Nat Genet. 613 2015;47(3):291-295. doi:10.1038/ng.3211 614 43. Jaholkowski P , Shadrin AA, Jangmo A, et al. Associations Between Symptoms of 615 Premenstrual Disorders and Polygenic Liability for Major Psychiatric Disorders. 616 JAMA Psychiatry. 2023;80(7):738-742. doi:10.1001/jamapsychiatry.2023.1137 617 44. Lin PC, Long CY , Ko CH, Yen JY. Comorbid Attention Deficit Hyperactivity 618 Disorder in Women with Premenstrual Dysphoric Disorder. J Womens Health 619 (Larchmt). 2024;33(9):1267-1275. doi:10.1089/jwh.2023.0907 620 45. Sepede G, Brunetti M, Di Giannantonio M. Comorbid Premenstrual Dysphoric 621 Disorder in Women with Bipolar Disorder: Management Challenges. 622 Neuropsychiatr Dis Treat. 2020;16:415-426. doi:10.2147/NDT.S202881 623 46. Kamijo S, Ishii Y, Horigane S ichiro, et al. A Critical Neurodevelopmental Role for 624 L-Type Voltage-Gated Calcium Channels in Neurite Extension and Radial 625 Migration. J Neurosci. 2018;38(24):5551-5566. doi:10.1523/JNEUROSCI.2357-626 17.2018 627 47. Pourtavakoli A, Ghafouri-Fard S. Calcium signaling in neurodevelopment and 628 pathophysiology of autism spectrum disorders. Mol Biol Rep. 2022;49(11):10811-629 10823. doi:10.1007/s11033-022-07775-6 630 48. Li HJ, Goff A, Rudzinskas SA, et al. Altered estradiol-dependent cellular Ca2+ 631 homeostasis and endoplasmic reticulum stress response in Premenstrual 632 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 26 Dysphoric Disorder. Mol Psychiatry. 2021;26(11):6963-6974. 633 doi:10.1038/s41380-021-01144-8 634 49. Zhang Y, Zhao Y, Song X, et al. Modulation of Stem Cells as Therapeutics for 635 Severe Mental Disorders and Cognitive Impairments. Front Psychiatry. 2020;11. 636 doi:10.3389/fpsyt.2020.00080 637 50. VanHouten JN, Neville MC, Wysolmerski JJ. The Calcium-Sensing Receptor 638 Regulates Plasma Membrane Calcium Adenosine Triphosphatase Isoform 2 639 Activity in Mammary Epithelial Cells: A Mechanism for Calcium-Regulated 640 Calcium Transport into Milk. Endocrinology. 2007;148(12):5943-5954. 641 doi:10.1210/en.2007-0850 642 51. Hantsoo L, Payne JL. Towards understanding the biology of premenstrual 643 dysphoric disorder: From genes to GABA. Neuroscience & Biobehavioral 644 Reviews. 2023;149:105168. doi:10.1016/j.neubiorev.2023.105168 645 52. Hantsoo L, Epperson CN. Allopregnanolone in premenstrual dysphoric disorder 646 (PMDD): Evidence for dysregulated sensitivity to GABA-A receptor modulating 647 neuroactive steroids across the menstrual cycle. Neurobiol Stress. 648 2020;12:100213. doi:10.1016/j.ynstr.2020.100213 649 53. Bigos KL, Mattay VS, Callicott JH, et al. Genetic variation in CACNA1C affects 650 brain circuitries related to mental illness. Arch Gen Psychiatry. 2010;67(9):939-651 945. doi:10.1001/archgenpsychiatry.2010.96 652 54. Harrison PJ, Husain SM, Lee H, et al. CACNA1C (CaV1.2) and other L-type 653 calcium channels in the pathophysiology and treatment of psychiatric disorders: 654 Advances from functional genomics and pharmacoepidemiology. 655 Neuropharmacology. 2022;220:109262. doi:10.1016/j.neuropharm.2022.109262 656 55. Dao DT, Mahon PB, Cai X, et al. Mood Disorder Susceptibility Gene CACNA1C 657 Modifies Mood-Related Behaviors in Mice and Interacts with Sex to Influence 658 Behavior in Mice and Diagnosis in Humans. Biological Psychiatry. 659 2010;68(9):801-810. doi:10.1016/j.biopsych.2010.06.019 660 56. Kempton MJ, Ruberto G, Vassos E, et al. Effects of the CACNA1C risk allele for 661 bipolar disorder on cerebral gray matter volume in healthy individuals. Am J 662 Psychiatry. 2009;166(12):1413-1414. doi:10.1176/appi.ajp.2009.09050680 663 57. Franke B, Vasquez AA, Veltman JA, Brunner HG, Rijpkema M, Fernández G. 664 Genetic variation in CACNA1C, a gene associated with bipolar disorder, 665 influences brainstem rather than gray matter volume in healthy individuals. Biol 666 Psychiatry. 2010;68(6):586-588. doi:10.1016/j.biopsych.2010.05.037 667 58. Coleman JRI, Gaspar HA, Bryois J, et al. The Genetics of the Mood Disorder 668 Spectrum: Genome-wide Association Analyses of More Than 185,000 Cases and 669 439,000 Controls. Biological Psychiatry. 2020;88(2):169-184. 670 doi:10.1016/j.biopsych.2019.10.015 671 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 27 59. Lee PH, Anttila V, Won H, et al. Genomic Relationships, Novel Loci, and 672 Pleiotropic Mechanisms across Eight Psychiatric Disorders. Cell. 673 2019;179(7):1469-1482.e11. doi:10.1016/j.cell.2019.11.020 674 60. Pardiñas AF, Holmans P, Pocklington AJ, et al. Common schizophrenia alleles 675 are enriched in mutation-intolerant genes and in regions under strong 676

selection. Nat Genet. 2018;50(3):381-389. doi:10.1038/s41588-018-677 0059-2 678 61. Identification of risk loci with shared effects on five major psychiatric disorders: a 679 genome-wide analysis. The Lancet. 2013;381(9875):1371-1379. 680 doi:10.1016/S0140-6736(12)62129-1 681 62. GWAS Catalog. Accessed September 25, 2024. 682 https://www.ebi.ac.uk/gwas/variants/rs758170 683 63. Bertone-Johnson ER, Hankinson SE, Willett WC, Johnson SR, Manson JE. 684 Adiposity and the development of premenstrual syndrome. J Womens Health 685 (Larchmt). 2010;19(11):1955-1962. doi:10.1089/jwh.2010.2128 686 64. Lu D, Aleknaviciute J, Kamperman AM, et al. Association Between Childhood 687 Body Size and Premenstrual Disorders in Young Adulthood. JAMA Network 688 Open. 2022;5(3):e221256. doi:10.1001/jamanetworkopen.2022.1256 689 65. Guintivano J, Byrne EM, Kiewa J, et al. Meta-Analyses of Genome-Wide 690 Association Studies for Postpartum Depression. AJP. 2023;180(12):884-895. 691 doi:10.1176/appi.ajp.20230053 692 66. Howard DM, Adams MJ, Clarke TK, et al. Genome-wide meta-analysis of 693 depression identifies 102 independent variants and highlights the importance of 694 the prefrontal brain regions. Nat Neurosci. 2019;22(3):343-352. 695 doi:10.1038/s41593-018-0326-7 696 67. Adams MJ, Streit F, Meng X, et al. Trans-ancestry genome-wide study of 697 depression identifies 697 associations implicating cell types and 698 pharmacotherapies. Cell. 2025;188(3):640-652.e9. 699 doi:10.1016/j.cell.2024.12.002 700 68. Li DJ, Tsai SJ, Bai YM, et al. Risks of major affective disorders following a 701 diagnosis of premenstrual dysphoric disorder: A nationwide longitudinal study. 702 Asian J Psychiatr. 2023;79:103355. doi:10.1016/j.ajp.2022.103355 703 69. Cirillo PC, Passos RBF, Bevilaqua MC do N, López JRRA, Nardi AE. Bipolar 704 disorder and Premenstrual Syndrome or Premenstrual Dysphoric Disorder 705 comorbidity: a systematic review. Braz J Psychiatry. 2012;34(4):467-479. 706 doi:10.1016/j.rbp.2012.04.010 707 70. Forrester-Knauss C, Zemp Stutz E, Weiss C, T schudin S. The interrelation 708 between premenstrual syndrome and major depression: Results from a 709 population-based sample. BMC Public Health. 2011;11(1):795. doi:10.1186/1471-710 2 458-11-795 711 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 28 71. Chen LC, Hsu JW, Huang KL, et al. Risk of developing major depression and 712 anxiety disorders among women with endometriosis: A longitudinal follow-up 713 study. J Affect Disord. 2016;190:282-285. doi:10.1016/j.jad.2015.10.030 714 72. Dubey N, Hoffman JF, Schuebel K, et al. The ESC/E(Z) complex, an effector of 715 response to ovarian steroids, manifests an intrinsic difference in cells from 716 women with premenstrual dysphoric disorder. Mol Psychiatry. 2017;22(8):1172-717 1184. doi:10.1038/mp.2016.229 718 73. Cohen LS, Soares CN, Otto MW, Sweeney BH, Liberman RF, Harlow BL. 719 Prevalence and predictors of premenstrual dysphoric disorder (PMDD) in older 720 premenopausal women: The Harvard Study of Moods and Cycles. Journal of 721 Affective Disorders. 2002;70(2):125-132. doi:10.1016/S0165-0327(01)00458-X 722 723 724 725 726 727 728 729 730 731 732 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 29 Figure 1. Manhattan plot from the meta-analyzed GWAS of DPS. This analysis included 17,511 cases and 50,633 controls. Dashed red line indicates genome-wide significant threshold P=5E-8. Pale dashed red line indicates suggestive significance threshold P=1.0x10-6. Green dots show SNPs in LD (r2>0.8) with the lead SNP of each genome-wide significant or suggestive locus. The point estimates of lead SNPs are provided in Table 2. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint Figure 2. Genetic correlations (rg) between DPS and psychosocial, gynecological traits, and steroid hormones. Dots indicate the point estimate while caps indicate the 95% confidence interval. Green dots denote empirical P value <0.05. The actual estimates are provided in Supplementary Table S4. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint Figure 3. Regional plot for the lead SNP rs758170. SNPs which are not in LD of the lead SNP in the selected region are colored grey. Only SNPs which are in LD of the lead SNP are displayed in the plot. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 32 Table 1. Characteristics of women with and without DPS in the LifeGene and MoBa cohorts. LifeGene MoBa Without DPS With DPS P-value Without DPS With DPS P-value Individuals, N 3267 1962 51,519 15,549 Age (years) 34 ±/i18.43 33 ± /i17.79 29.47 ± 4.56 30.1 ± 4.68 Educational Levela 0.060 <0.001 Pre-secondary education 112 (3.42) 48 (2.5) 1137 (2.2) 427 (2.7) Secondary education 777 (23.7) 513 (26.1) 15815 (30.7) 5393 (34.7) Post-secondary education 2378 (72.4) 1401 (71.1) 31936 (62.0) 8850 (56.9) Unknown 12 (0.4) 5 (0.3) 2631 (5.1) 879 (5.7) Civil Statusa 0.024 0.003 Single/separated/widower 1922 (58.8) 1228 (62.5) 915 (1.8) 301 (1.9) Married/cohabitant 1342 (41.0) 733 (37.3) 46648 (90.5) 13937 (89.6) Unknown 3 (0.09) 1 (0.05) 3956 (7.7) 1311 (8.4) Depressionb <0.001 <0.001 No 2795 (85.6) 1390 (70.9) 47111 (91.4) 13208 (84.9) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 33 LifeGene MoBa Without DPS With DPS P-value Without DPS With DPS P-value Yes 472 (14.4) 572 (29.1) 4408 (8.6) 2341 (15.1) Anxiety disorderb <0.001 <0.001 No 2738 (83.8) 1316 (67.1) 47895 (93.0) 13882 (89.3) Yes 529 (16.2) 646 (32.9) 3624 (7) 1667 (10.7) Age at menarche <= 10 years 13.39 ± 3.64 82 (2.5) 13.20 ± 3.50 81 (4.1) 0.005 13.05 ± 1.37 970 (1.9) 12.93 ± 1.39 387 (2.5) = 15 years 419 (12.8) 232 (11.9) 6860 (13.3) 1807 (11.6) Unknown 189 (5.8) 97 (4.9) 653 (1.3) 183 (1.2) Paritya 0.134 =3 231 (7.1) 152 (7.8) 1243 (2.4) 568 (3.7) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 34 sd standard deviation, PND perinatal depression mean/i1±/i1sd or N (%). MoBa Norwegian Mother, Father and Child Cohort Study. P-values were obtained from chi-squared test. a At baseline. b Lifetime (in LifeGene). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 35 Table 2. Lead SNPs associated with DPS: a meta-analysis of the LifeGene and the MoBa. Lead SNPs are independent, significant SNPs that have reached either GWAS significant P-value <5x10-8 or borderline significance P-value <5x10-6 in the genomic rick loci. The SNPs are presented in ascending order or the P-values. Chr Position rsID Gene A1 A2 Sample OR (95% CI) a P-value P for heterogeneity 12 2361460 rs758170 CACNA1C C T LifeGene 0.95 (0.86– 1.04) 0.242 MoBa 0.92 (0.90– 0.95) 2.68 x 10-8 Meta-analysis 0.93 (0.90– 0.95) 1.53 x 10-8 0.5949 5 144181270 rs76665457 CTB-85P21.1 G C LifeGene 0.80 (0.63– 0.99) 0.032 MoBa 0.86 (0.81– 0.91) 5.30 x 10-7 Meta-analysis 0.86 (0.80– 0.86) 5.96 x 10-8 0.5668 10 56883744 rs12770903 PCDH15 G T LifeGene 0.81 (0.62– 1.02) 0.066 MoBa 0.85 (0.80– 0.91) 7.31 x 10-7 Meta-analysis 0.84 (0.78– 0.90) 1.34 x 10-7 0.7665 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 36 Chr Position rsID Gene A1 A2 Sample OR (95% CI) a P-value P for heterogeneity 4 15764240 rs147346386 CD38 C T LifeGene 0.60 (0.39– 0.85) 0.001 MoBa 0.84 (0.78– 0.90) 5.89 x 10-6 Meta-analysis 0.82 (0.77– 0.89) 1.95 x 10-7 0.0813 13 90887997 rs4773561 KRT18P27 G A LifeGene 0.89 (0.80– 0.98) 0.013 MoBa 0.93 (0.91– 0.96) 2.96 x 10-6 Meta-analysis 0.93 (0.90– 0.95) 2.06 x 10-7 0.3235 1 151581008 rs77519409 RP11-404E16.1 G A LifeGene - - MoBa 1.17 (1.10– 1.25) 4.61 x 10-7 Meta- analysis 1.17 (1.10– 1.25) 4.61 x 10-7 - 18 77629986 rs112526506 KCNG2 G A LifeGene 1.08 (0.95– 1.23) 0.183 MoBa 1.10 (1.06– 1.50 x 10-6 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint 37 Chr Position rsID Gene A1 A2 Sample OR (95% CI) a P-value P for heterogeneity 1.15) Meta- analysis 1.10 (1.06– 1.14) 6.12 x 10-7 0.8539 Chr chromosome, Position on the chromosome (GRCh37 genomic build), SNP single-nucleotide polymorphism, A1 Risk allele, A2 reference allele, OR odds ratio, CI confidence interval, P-value, MoBa Norwegian Mother, Father and Child Cohort Study. aEstimates were adjusted for 10 principal components in LifeGene and MoBa; and were additionally adjusted for the sub-study membership and year of birth, respectively. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted December 2, 2025. ; https://doi.org/10.64898/2025.12.01.25341208doi: medRxiv preprint