Folate intake, MTHFR genotype and premenstrual symptoms.

Participants were from the Toronto Nutrigenomics and Health (TNH) study, which is a cross-sectional study examining the effects of genetic variation and dietary intake on biomarkers of health. The TNH study consists of a multiethnic population of young adults and recruitment occurred between 2004 and 2010 at the University of Toronto campus. Participants who met the inclusion criteria provided written informed consent, and the study protocol was approved by the University of Toronto’s Research Ethics Board. Women who were pregnant or breastfeeding were not eligible to participate. For the present analyses, male participants ( n 532) were excluded ( Figure 1 ). Those who reported a diagnosis of endometriosis, amenorrhea or polycystic ovarian syndrome ( n 20), smokers ( n 52), users of hormonal contraceptives ( n 304) and users of anxiolytics or anti-depressants ( n 24) were also excluded due to the potential confounding effects of these variables on premenstrual symptoms or folate status. Those with missing data from the General Health and Lifestyle Questionnaire (GHLQ), or missing genetic data were also excluded ( n 3). Individuals who may have underreported ( 14644 kJ/day (3500 kcal/d)) their energy intakes were excluded ( n 36). Based on their self-reported ethnocultural status, the remaining 678 participants were divided into four major ethnic groups: Caucasian ( n 243), East Asian ( n 212), South Asian ( n 78) and Other ( n 45). Caucasians were those who identified as European, Middle Eastern, or Hispanic. East Asians included people who self-reported from China, Korea, Japan, Philippines, Vietnam, Thailand and Cambodia. South Asians consisted of Bangladeshis, Sri Lankans, Indians and Pakistanis. Aboriginal Canadians, Afro-Caribbeans and those who self-reported belonging to two or more ethnic groups were included in the Other category. Figure 1. Figure 1 long description. Participant flowchart summarising exclusion criteria for the present study. TNH, Toronto Nutrigenomics and Health Study. A flowchart illustrating the exclusion criteria for participants in the Toronto Nutrigenomics and Health Study, starting with 1649 participants and ending with a total sample size of 678. Figure 1 long description. Participant flowchart summarising exclusion criteria for the present study. TNH, Toronto Nutrigenomics and Health Study. A GHLQ, which consisted of a comprehensive set of questions examining age, sex, ethnocultural group, current medical conditions, medication use, hormonal contraceptive use, dietary supplements, special diets and physical activity levels, was completed by all participants ( 21 , 22 ) . The GHLQ contained a premenstrual symptoms questionnaire, as previously described ( 23 ) , that assessed the severity and presence of fifteen commonly reported premenstrual symptoms within 5 d of the onset of menstruation and up to 4 d afterwards. There were four severity levels: ‘none’, ‘mild’, ‘moderate’ and ‘severe’ that were self-reported by all female participants. This questionnaire was developed by combining the most frequently reported symptoms in the literature and previously validated questionnaires ( 24 , 25 ) . The questionnaire was used in several previous studies ( 1 , 21 – 23 , 26 , 27 ) . Recent research on mental health conditions supports a shift from broad syndromic classifications to more specific symptoms and subtypes ( 28 ) , reinforcing our methodological choice of assessing symptoms individually. Given the heterogeneity of PMS classification, analysing individual symptoms allows for a more precise understanding of their distinct biological, hormonal and genetic mechanisms. As each woman experiences a unique combination of symptoms, this approach better captures the complexity of premenstrual symptoms rather than assuming a uniform aetiology for multiple symptoms as a single condition. Trained personnel took anthropometric measurements such as weight, height and waist circumference while participants wore light clothing and had their shoes removed. BMI was calculated using these measurements (weight(kg)/height(m) 2 ). The GHLQ also included self-reported physical activity, with participants estimating their own activity levels, which included time spent sleeping and engaging in light to vigorous activity. These values were then converted to weekly metabolic hour equivalents (met-h/wk). A validated 196-item Toronto-modified Harvard FFQ was used to assess habitual intake over the previous month, including foods, beverages and supplements ( 29 ) . As previously described, participants’ responses to each individual item were used to calculate nutrient intakes using the nutrient contents of the food in the US Department of Agriculture’s database ( 30 ) . For the analyses presented here, folate intakes from food and supplements were calculated using data from the FFQ. By selecting from several frequency options, subjects estimated their consumption of a predetermined portion of each item over the previous month. Following that, responses were converted into estimated daily averages of total folate intake from foods and supplements. All participants provided an overnight 12-hour fasting blood sample, and DNA was extracted from whole blood cells using previously described methods at LifeLabs medical laboratory services (Toronto, Canada) ( 31 ) . DNA was extracted from white blood cells in the peripheral circulation. MTHFR (rs1801133) genotyping (C677T) was performed on all participants using Sequenom MassARRAY® technology, as previously described ( 32 ) at the Clinical Genomics Centre in Princess Margaret Hospital (Toronto, Canada). R and RStudio were used for all statistical analyses (version 4.2.0). The distribution of continuous variables was evaluated for normality prior to analysis. BMI was not normally distributed and was thus log-transformed for all analysis. The α was set at 0·05, and all P -values reported are two-sided. In this study, severities were classified as ‘present’ (mild, moderate or severe) or ‘absent’ (none) to examine the prevalence of premenstrual symptoms. Subject characteristics were compared between participants with above and below the median folate intake (647 mcg/d) using analysis of variance for continuous variables and chi-square tests for categorical variables. The recommended daily allowance of folate intake for females over 19 years of age is 400 mcg/d, and the upper limit is 1000 mcg/d ( 33 ) . Binomial logistic regressions were used to determine the association between folate intake and MTHFR genotypes and premenstrual symptoms, separately. Logistic regressions were also employed to determine whether MTHFR genotype modified the association between folate intake and the presence of premenstrual symptoms. OR and corresponding 95 % CI were calculated to compare the odds of experiencing premenstrual symptoms for participants with low v . high folate intake (i.e. below and above the median of 647 mcg/d) using ‘no symptoms’ as the reference category, and between MTHFR genotypes, using the CC genotype as the reference category. OR and 95 % CI were also calculated to compare the odds of experiencing premenstrual symptoms in low v . high folate intake stratified by MTHFR genotype. For each symptom, the P -values for the interaction term between MTHFR and folate intake were calculated. Univariate and multivariate analyses were performed with adjustments for age, ethnicity, log-transformed BMI, physical activity, total vitamin B 12 intake and energy intake. The interaction term P -values were subjected to Benjamini–Yekutieli adjustments for multiple comparisons (15 tests, α = 0·05: P < 0·015). Although genetic variants typically have a small effect size on the outcome observed, the MTHFR genotype is strongly associated with folate status ( 10 – 12 ) . Thus, the statistical analyses performed in this study were adequately powered, even with a small effect size of 0·2 ( 34 ) . Sample size was determined using a power of 0·8, a small effect size of 0·2 and a significance of P < 0·05 using R (Version 4.0.3) and RStudio (version 1.3.1). Based on this equation, the observed sample sizes in previous studies assessing MTHFR and other outcomes ( 13 , 35 , 36 ) , an adequate sample size is estimated to be approximately 190 participants.

Participant characteristics were stratified by MTHFR genotype and are presented in Table 1 . Most participants had the CC genotype (47 %), followed by the CT (42 %) and TT (11 %) genotypes. There were no significant differences between genotypes for age, BMI, physical activity level, total folate intake and total vitamin B 12 intake. Energy intake was significantly higher among those with the TT genotype, compared with the CT and CC genotypes ( P = 0·03). Protein and total fat intake were also significantly higher among those with the TT genotype in comparison to the CT and CC genotype carriers ( P = 0·008 and 0·05, respectively). Table 1. Participant characteristics stratified by MTHFR genotype * (Numbers and percentages; mean values and standard deviations) Table 1 long description. CC CT TT P † n % n % n % n , (%) 316 47 286 42 76 11 Mean sd Mean sd Mean sd Age, (years) 22·4 2·5 22·4 2·4 22·33 2·59 0·94 n % n % n % Ethnicity, n (%) 0·001  Caucasian 96 40 115 47 32 13  East Asian 128 41 148 47 36 11  South Asian 59 76 15 19 4 5  Other 33 73 8 18 4 9 45 Mean sd Mean sd Mean sd BMI (kg/m 2 ) 22·42 3·68 22·14 3·14 21·41 2·82 0·06 Physical activity level (MET-h/week) ‡ 1·74 0·39 1·73 0·44 1·67 0·37 0·38 Energy (kcal/d) 1781 609 1884 622 1953 673 0·03 Carbohydrates (g/d) 240 92 247 90 260 98 0·21 Total fat (g/d) 58 24 64 27 65 27 0·008 Protein (g/d) 78 29 83 33 84 31 0·05 Thiamin (mg/d) 2·3 5·1 2·7 6·0 2·0 1·1 0·58 Riboflavin (mg/d) 2·9 5·2 3·2 6·1 2·5 1·3 0·53 Niacin (mg/d) 26·2 13·6 27·9 18·7 27·4 14 0·42 Pantothenic acid (mg/d) 7·8 6·8 8·6 8·5 7·4 3·7 0·30 Vitamin B 6 (mg/d) 4·1 13·2 4·7 14·4 4·9 17·8 0·84 Folate (mcg/d) 780 488 776 464 755 420 0·91 Vitamin B 12 (mg/d) 7·4 7·4 8·4 9·3 6·7 3·9 0·14 A table showing participant characteristics stratified by MTHFR genotype, including age, ethnicity, BMI, physical activity level, and nutrient intake. * Values are unadjusted means (standard deviations) for continuous variables unless otherwise indicated. † Differences between groups were compared using chi-square tests for categorical variables and ANOVA for continuous variables. ‡ MET: metabolic equivalent. Participant characteristics stratified by MTHFR genotype * (Numbers and percentages; mean values and standard deviations) Table 1 long description. Values are unadjusted means (standard deviations) for continuous variables unless otherwise indicated. Differences between groups were compared using chi-square tests for categorical variables and ANOVA for continuous variables. MET: metabolic equivalent. Table 2 presents the associations among total folate intake, stratified by the median (647 mcg/d) and individual premenstrual symptoms. There were no observed associations between folate intake and any premenstrual symptoms, in both univariate and multivariate models, adjusting for age, log-transformed BMI, ethnicity, physical activity level, energy intake and total vitamin B 12 intake. Table 2. Associations between higher folate intake and risk of premenstrual symptoms (Numbers and percentages; OR and 95 % CI) Table 2 long description. Premenstrual symptom, n (%) Folate intake < 651 mcg/d ( n 337) Folate intake ≥ 651 mcg/d ( n 341) Unadjusted OR 95 CI Unadjusted P Adjusted OR 95 CI * Adjusted P * Acne/skin blemish  No 146 (43 %) 129 (38 %) REF REF  Yes 191 (57 %) 212 (62 %) 0·8 (0·59, 1·08) 0·15 0·83 (0·58, 1·18) 0·3 Desire to be alone  No 218 (65 %) 216 (63 %) REF REF  Yes 119 (35 %) 125 (37 %) 0·94 (0·69, 1·29) 0·72 0·92 (0·64, 1·32) 0·66 Anxiety/tension/nervousness  No 210 (62 %) 216 (64 %) REF REF  Yes 127 (38 %) 125 (36 %) 1·05 (0·76, 1·43) 0·78 1·26 (0·88, 1·8) 0·21 Increased appetite/food cravings  No 130 (39 %) 124 (36 %) REF REF  Yes 207 (61 %) 217 (64 %) 0·91 (0·67, 1·24) 0·55 0·87 (0·61, 1·25) 0·46 Bloating/swelling/breast tenderness  No 97 (29 %) 89 (26 %) REF REF  Yes 240 (71 %) 252 (74 %) 0·87 (0·62, 1·23) 0·43 0·9 (0·61, 1·33) 0·61 Clumsiness  No 286 (85 %) 286 (84 %) REF REF  Yes 51 (15 %) 55 (16 %) 0·93 (0·61, 1·4) 0·72 1·19 (0·74, 1·94) 0·47 Confusion/difficulty concentrating/forgetfulness  No 261 (78 %) 263 (77 %) REF REF  Yes 76 (23 %) 78 (23 %) 0·98 (0·69, 1·41) 0·92 1·12 (0·74, 1·7) 0·6 Cramps  No 80 (24 %) 76 (22 %) REF REF  Yes 257 (76 %) 265 (8 %) 0·92 (0·64, 1·32) 0·65 0·98 (0·65, 1·49) 0·94 Depression  No 245 (73 %) 239 (70 %) REF REF  Yes 92 (27 %) 102 (30 %) 0·88 (0·63, 1·23) 0·45 0·98 (0·67, 1·44) 0·92 Fatigue  No 145 (43 %) 157 (46 %) REF REF  Yes 192 (57 %) 184 (54 %) 1·13 (0·83, 1·53) 0·43 1·02 (0·72, 1·45) 0·91 Headaches  No 257 (76 %) 243 (71 %) REF REF  Yes 80 (24 %) 98 (29 %) 0·77 (0·55, 1·09) 0·14 0·78 (0·52, 1·16) 0·22 Insomnia  No 298 (88 %) 303 (89 %) REF REF  Yes 39 (12 %) 38 (11 %) 1·04 (0·65, 1·68) 0·86 1·19 (0·69, 2·06) 0·53 Mood swings/crying easily/irritability/angry outbursts  No 94 (28 %) 92 (27 %) REF REF  Yes 243 (72 %) 249 (73 %) 0·96 (0·68, 1·34) 0·79 1·08 (0·73, 1·59) 0·7 Nausea  No 288 (85 %) 291 (85 %) REF REF  Yes 49 (15 %) 50 (15 %) 0·99 (0·65, 1·52) 0·96 1·08 (0·66, 1·78) 0·77 Sexual desire/activity change  No 188 (56 %) 166 (49 %) REF REF  Yes 149 (44 %) 175 (51 %) 0·75 (0·56, 1·02) 0·06 0·81 (0·57, 1·15) 0·25 Table showing associations between folate intake and premenstrual symptoms with data on percentages and odds ratios. * P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake ( 3500 kcal/d), and individuals with missing information. Associations between higher folate intake and risk of premenstrual symptoms (Numbers and percentages; OR and 95 % CI) Table 2 long description. P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake ( 3500 kcal/d), and individuals with missing information. The associations between total folate intake and individual premenstrual symptoms stratified by MTHFR genotype are shown in Table 3 . There were no associations between folate intake and any premenstrual symptoms when stratified by MTHFR genotype. Moreover, MTHFR did not modify the association between folate intake and premenstrual symptoms after adjusting for covariates, as shown in Table 4 . Table 3. Associations between higher folate (> 647 mcg/d) intake and premenstrual symptoms stratified by MTHFR genotype (OR and 95 % CI) Table 3 long description. Premenstrual Symptom Unadjusted Unadjusted Adjusted Adjusted Unadjusted Unadjusted Adjusted Adjusted Unadjusted Unadjusted Adjusted Adjusted OR 95 % CI P OR 95 % CI * P * OR 95 % CI P OR 95 % CI * P * OR 95 % CI P OR 95 % CI * P * CC CT TT Acne/skin blemish 0·76 0·49, 1·19 0·23 0·8 0·49, 1·3 0·37 0·8 0·5, 1·28 0·35 0·82 0·49, 1·37 0·45 1·03 0·4, 2·65 0·97 0·99 0·37, 2·61 0·98 Desire to be alone 0·79 0·49, 1·26 0·32 0·74 0·44, 1·22 0·23 1·08 0·67, 1·74 0·74 1·04 0·62, 1·75 0·88 1·22 0·47, 3·18 0·68 1·37 0·51, 3·67 0·53 Anxiety/tension/nervousness 1·23 0·77, 1·95 0·38 1·4 0·85, 2·31 0·18 0·88 0·55, 1·42 0·60 1·09 0·65, 1·83 0·75 1·09 0·43, 2·76 0·86 1·37 0·52, 3·58 0·52 Increased appetite/food changes 0·92 0·58, 1·45 0·71 0·88 0·54, 1·44 0·61 0·9 0·56, 1·45 0·66 0·85 0·5, 1·43 0·53 0·93 0·35, 2·44 0·87 0·92 0·34, 2·51 0·88 Bloating/swelling/breast tenderness 0·9 0·56, 1·44 0·65 0·94 0·56, 1·57 0·81 0·94 0·53, 1·65 0·82 0·88 0·48, 1·63 0·69 0·73 0·27, 1·91 0·52 0·83 0·31, 2·27 0·72 Clumsiness 0·98 0·52, 1·84 0·95 1·18 0·6, 2·34 0·63 0·88 0·48, 1·6 0·67 1·18 0·61, 2·29 0·62 1·06 0·25, 4·59 0·94 1·38 0·31, 6·24 0·67 Confusion/difficulty concentrating/forgetfulness 0·72 0·4, 1·27 0·26 0·73 0·4, 1·36 0·33 1·2 0·71, 2·04 0·49 1·49 0·83, 2·66 0·18 1·41 0·51, 3·93 0·51 1·55 0·53, 4·52 0·42 Cramps 0·65 0·37, 1·13 0·13 0·75 0·41, 1·36 0·34 1·35 0·79, 2·33 0·27 1·39 0·77, 2·52 0·27 0·71 0·25, 1·98 0·51 0·78 0·27, 2·25 0·64 Depression 0·68 0·41, 1·13 0·14 0·72 0·41, 1·25 0·24 1·13 0·68, 1·87 0·64 1·24 0·71, 2·15 0·45 0·97 0·39, 2·46 0·96 1·26 0·48, 3·28 0·64 Fatigue 1·18 0·75, 1·84 0·47 1·06 0·65, 1·71 0·82 1·17 0·74, 1·87 0·50 1·01 0·61, 1·68 0·96 0·81 0·33, 2 0·65 0·91 0·36, 2·31 0·85 Headaches 0·78 0·48, 1·28 0·33 0·75 0·44, 1·28 0·29 0·82 0·47, 1·42 0·48 0·87 0·48, 1·59 0·65 0·52 0·18, 1·53 0·24 0·65 0·22, 1·97 0·45 Insomnia 0·93 0·46, 1·9 0·85 1·06 0·49, 2·29 0·88 1·44 0·7, 2·96 0·32 1·7 0·77, 3·76 0·19 0·49 0·11, 2·1 0·33 0·5 0·11, 2·27 0·37 Mood swings/crying easily/irritability/angry outbursts 1·06 0·65, 1·72 0·81 1·18 0·7, 1·99 0·54 0·86 0·51, 1·46 0·58 0·99 0·56, 1·75 0·97 0·93 0·34, 2·59 0·89 1·01 0·36, 2·9 0·98 Nausea 1·07 0·58, 1·97 0·82 1·15 0·59, 2·26 0·68 1·03 0·51, 2·09 0·93 1·08 0·5, 2·33 0·85 0·65 0·2, 2·03 0·45 0·96 0·29, 3·2 0·95 Sexual desire/activity change 0·98 0·63, 1·52 0·91 1·09 0·67, 1·76 0·74 0·58 0·36, 0·93 0·02 0·57 0·34, 0·96 0·03 0·64 0·26, 1·6 0·34 0·88 0·34, 2·27 0·78 A table showing the associations between total folate intake and individual premenstrual symptoms stratified by MTHFR genotype. * P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: smokers, hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, implausible energy intake ( 3500 kcal/d) and individuals with missing information. Ref = ‘no’ premenstrual symptoms (removed for space). Associations between higher folate (> 647 mcg/d) intake and premenstrual symptoms stratified by MTHFR genotype (OR and 95 % CI) Table 3 long description. P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: smokers, hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, implausible energy intake ( 3500 kcal/d) and individuals with missing information. Ref = ‘no’ premenstrual symptoms (removed for space). Table 4. Interaction between MTHFR and folate intake premenstrual symptoms Table 4 long description. Premenstrual symptom MTHFR and folate intake interaction * * Acne/skin blemish 0·92 Desire to be alone 0·42 Anxiety/tension/nervousness 0·75 Increased appetite/food cravings 0·98 Bloating/swelling/breast tenderness 0·97 Clumsiness 0·98 Confusion/difficulty concentrating/forgetfulness 0·18 Cramps 0·28 Depression 0·29 Fatigue 0·96 Headaches 0·87 Insomnia 0·89 Mood swings/ crying easily/irritability/angry outbursts 0·31 Nausea 0·96 Sexual desire/activity change 0·16 Table showing interaction between MTHFR and folate intake on premenstrual symptoms with 15 rows and 2 columns. * P -values for the interaction between MTHFR and folate intake were adjusted for age, ethnicity, log-transformed BMI, physical activity levels, energy intake and vitamin B 12 intake. Interaction between MTHFR and folate intake premenstrual symptoms Table 4 long description. P -values for the interaction between MTHFR and folate intake were adjusted for age, ethnicity, log-transformed BMI, physical activity levels, energy intake and vitamin B 12 intake. Table 5 shows the association of MTHFR genotype in those with folate intake below the median (< 647 mcg/d) and premenstrual symptoms. In these analyses, those with the CT genotype had increased odds of reporting premenstrual confusion/difficulty concentrating/forgetfulness (OR: 2·19; 95 % CI: 1·16, 3·80), compared with those with the CC genotype. Those with the CT genotype had higher odds of experiencing premenstrual depression, compared with those with the CC genotype (OR: 1·67; 95 % CI: 1·00, 2·82) in the univariate analyses – this finding was no longer significant when adjusted for covariates. Those with the TT genotype were at a higher risk of experiencing premenstrual depression (OR: 2·41; 95 % CI: 1·08, 5·38), after adjusting for covariates compared with those with the CC genotype. There were no further associations observed between other premenstrual symptoms and MTHFR genotype in women with lower folate intake. Table 5. Associations between MTHFR genotype and premenstrual symptoms in those with lower folate intake (< 647 mcg/d) (OR and 95 % CI) Table 5 long description. Premenstrual symptom MTHFR Genotype Unadjusted Unadjusted Adjusted Adjusted OR 95 % CI P value OR 95 % CI * P value * Acne/skin blemish CC CT 1·23 0·77, 1·94 0·38 1·02 0·63, 1·65 0·95 TT 1·62 0·77, 3·40 0·20 1·30 0·60, 2·85 0·50 Desire to be alone CC CT 1·47 0·91, 2·37 0·11 1·50 0·96, 2·64 0·07 TT 1·19 0·56, 2·53 0·65 1·50 0·68, 3·30 0·31 Anxiety/tension/nervousness CC CT 0·97 0·60, 1·55 0·88 0·90 0·55, 1·48 0·68 TT 0·99 0·47, 2·08 0·98 0·95 0·44, 2·07 0·91 Increased appetite/food cravings CC CT 0·99 0·62, 1·59 0·97 0·97 0·59, 1·59 0·90 TT 1·35 0·63, 2·79 0·44 1·52 0·69, 3·38 0·30 Bloating/swelling/breast tenderness CC CT 1·77 1·05, 2·98 0·03 1·75 1·02, 3·00 0·04 TT 0·91 0·43, 1·94 0·81 0·95 0·44, 2·09 0·91 Clumsiness CC CT 1·29 0·69, 2·42 0·49 1·30 0·67, 2·50 0·44 TT 0·74 0·24, 2·29 0·60 0·68 0·21, 2·22 0·52 Confusion/difficulty concentrating/forgetfulness CC CT 2·23 1·19, 3·68 0·009 2·19 1·16, 3·80 0·01 TT 2·22 0·98, 5·07 0·06 2·02 0·84, 4·88 0·11 Cramps CC CT 1·12 0·65, 1·93 0·67 1·19 0·67, 2·10 0·55 TT 0·74 0·34, 1·64 0·46 0·76 0·33, 1·77 0·52 Depression CC CT 1·67 1·00, 2·82 0·05 1·66 0·96, 2·87 0·06 TT 2·22 1·04, 4·78 0·03 2·41 1·08, 5·38 0·03 Fatigue CC CT 0·97 0·61, 1·55 0·91 1·14 0·70, 1·86 0·60 TT 0·68 0·33, 1·39 0·28 0·85 0·40, 1·80 0·67 Headaches CC CT 0·78 0·48, 1·34 0·36 0·97 0·55, 1·72 0·92 TT 0·65 0·27, 1·60 0·35 0·97 0·38, 2·48 0·94 Insomnia CC CT 1·38 0·68, 2·78 0·36 1·59 0·75, 3·38 0·23 TT 0·76 0·21, 2·75 0·67 0·85 0·22, 2·24 0·81 Mood swings/crying easily/irritability/angry outbursts CC CT 1·02 0·62, 1·70 0·93 0·98 0·59, 1·68 0·96 TT 1·06 0·47, 2·37 0·88 0·91 0·39, 2·10 0·82 Nausea CC CT 0·75 0·38, 1·45 0·38 0·71 0·36, 1·42 0·33 TT 1·02 0·38, 2·70 0·97 1·14 0·41, 3·17 0·80 Sexual desire/activity change CC CT 0·70 0·44, 1·10 0·12 0·63 0·39, 1·03 0·06 TT 0·63 0·30, 1·32 0·22 0·61 0·28, 1·33 0·22 A table showing the association of MTHFR genotype with premenstrual symptoms in individuals with folate intake below the median. * P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: individuals with folate levels ≥ 647 mcg/d, smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake ( 3500 kcal/d) and individuals with missing information. Associations between MTHFR genotype and premenstrual symptoms in those with lower folate intake (< 647 mcg/d) (OR and 95 % CI) Table 5 long description. P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: individuals with folate levels ≥ 647 mcg/d, smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake ( 3500 kcal/d) and individuals with missing information. Similar genetic association analyses were carried out between MTHFR genotype and individual premenstrual symptoms, but in those above the median folate intake (≥ 647 mcg/d) and are presented in Table 6 . In this analysis, we found that women with higher folate intake, who have the CT genotype of MTHFR , had increased odds of reporting premenstrual bloating/swelling/breast tenderness, compared with those with the CC genotype after adjusting for covariates (OR: 1·93; 95 % CI: 1·09, 3·21). Women who consumed higher folate, and who had the CT genotype, had lower odds of experiencing premenstrual cramps, compared with those with CC genotype (OR: 0·54; 95 % CI: 0·31, 0·94). However, this association was no longer observed after adjusting for covariates. Table 6. Associations between MTHFR genotype and premenstrual symptoms in those with higher folate intake (≥ 647 mcg/d) (OR and 95 % CI) Table 6 long description. Premenstrual symptom MTHFR genotype Unadjusted Unadjusted Adjusted Adjusted OR 95 % CI P value OR 95 % CI * P value * Acne/skin blemish CC CT 1·17 0·73, 1·86 0·52 1·03 0·63, 1·70 0·90 TT 1·18 0·57, 2·46 0·65 1·15 0·53, 2·48 0·72 Desire to be alone CC CT 1·07 0·67, 1·71 0·77 1·16 0·70, 1·92 0·57 TT 0·77 0·36, 1·64 0·50 0·85 0·39, 1·87 0·69 Anxiety/tension/nervousness CC CT 1·35 0·84, 2·16 0·21 1·24 0·75, 2·05 0·41 TT 1·12 0·55, 2·34 0·76 1·11 0·51, 2·39 0·79 Increased appetite/food cravings CC CT 1·01 0·64, 1·62 0·95 1·02 0·61, 1·69 0·95 TT 1·33 0·63, 2·85 0·45 1·47 0·66, 3·25 0·34 Bloating/swelling/breast tenderness CC CT 1·69 1·00, 2·86 0·05 1·93 1·09, 3·21 0·02 TT 1·13 0·52, 2·46 0·76 1·12 0·50, 2·54 0·77 Clumsiness CC CT 1·44 0·78, 2·65 0·24 1·30 0·66, 2·51 0·44 TT 0·68 0·22, 2·11 0·50 0·62 0·19, 2·01 0·43 Confusion/difficulty concentrating/forgetfulness CC CT 1·25 0·73, 2·14 0·42 1·28 0·71, 2·29 0·40 TT 1·13 0·49, 2·64 0·77 1·29 0·53, 3·12 0·57 Cramps CC CT 0·54 0·31, 0·94 0·02 0·66 0·36, 1·20 0·17 TT 0·68 0·29, 1·61 0·38 0·70 0·29, 1·73 0·44 Depression CC CT 1·00 0·61, 1·66 0·98 1·08 0·63, 1·86 0·77 TT 1·55 0·74, 3·23 0·24 1·46 0·67, 3·18 0·34 Fatigue CC CT 0·98 0·63, 1·54 0·92 1·15 0·70, 1·88 0·57 TT 0·98 0·48, 1·20 0·96 0·92 0·44, 1·94 0·82 Headaches CC CT 0·75 0·45, 1·24 0·26 0·77 0·45, 1·33 0·35 TT 0·97 0·45, 2·08 0·94 1·01 0·45, 2·25 0·98 Insomnia CC CT 0·90 0·43, 1·87 0·77 0·95 0·42, 2·14 0·90 TT 1·45 0·53, 3·98 0·47 1·86 0·63, 5·50 0·26 Mood swings/crying easily/irritability/angry outbursts CC CT 1·26 0·76, 2·10 0·37 1·12 0·65, 1·96 0·68 TT 1·21 0·54, 2·70 0·64 1·16 0·50, 2·68 0·73 Nausea CC CT 0·78 0·40, 1·51 0·45 0·84 0·41, 1·73 0·64 TT 1·70 0·71, 4·05 0·23 1·54 0·60, 3·95 0·36 Sexual desire/activity change CC CT 1·18 0·75, 1·85 0·49 1·30 0·80, 2·14 0·30 TT 0·96 0·47, 1·95 0·92 0·87 0·41, 1·84 0·72 Table showing associations between MTHFR genotype and premenstrual symptoms in women with higher folate intake. * P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: individuals with folate levels < 647 mcg/d, smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake (3500 kcal/d) and individuals with missing information. Associations between MTHFR genotype and premenstrual symptoms in those with higher folate intake (≥ 647 mcg/d) (OR and 95 % CI) Table 6 long description. P value adjusted for age, log-transformed BMI, ethnicity, physical activity level, energy intake and vitamin B 12 intake. Excluded: individuals with folate levels < 647 mcg/d, smokers, those with a hormonal imbalance diagnosis, HC users, those on antidepressants or anxiolytics, those with an implausible energy intake (3500 kcal/d) and individuals with missing information.

The purpose of this study was to examine the association between folate, using dietary measures and genetic variation in MTHFR , and premenstrual symptoms. We observed that in women with lower folate intake, carriers of the Tallele had increased odds of reporting premenstrual depression, compared with those with the CC genotype. To our knowledge, this is the first study to investigate the role of MTHFR in premenstrual symptoms. Variations in the MTHFR gene determine the way individuals use dietary folate. The most commonly researched genetic variant affecting folate metabolism, a variant in the MTHFR gene (rs1801133) involving a cytosine to thymidine (C → T) transition at nucleotide 677, is associated with reduced enzyme activity, DNA hypomethylation, developmental anomalies and elevated plasma homocysteine levels, which have been shown to increase the risk of cardiovascular disease ( 35 ) . Individuals with the TT genotype for the C677T polymorphism have increased plasma homocysteine and lower serum folate concentrations compared with those who have the CC genotype and heterozygotes appear to have an intermediate phenotype ( 11 , 13 , 15 , 35 ) . Many polymorphisms in genes involved in folate and homocysteine metabolism have been examined for their effects on folate and homocysteine levels, but the most consistent, and the largest effect is observed with the MTHFR C677T variant ( 37 , 38 ) . Folate is involved in neurotransmitter synthesis directly through its biologically active form, L-methyl-folate ( 39 ) . This active form of folate is also involved in converting homocysteine to methionine, which then converts to S-adenosyl-methionine, which is an important methyl donor in the synthesis of norepinephrine, dopamine and serotonin – all of which are neurotransmitters that play a role in depression ( 40 , 41 ) . Our research is consistent with previous findings on folate and depression ( 18 , 39 ) and further extends this relationship to our understanding of premenstrual depression and folate. A meta-analysis of randomised control trials examined the efficacy of adjunctive folate for major depressive disorders and suggested that folate was superior to placebo in treating depressive disorders and other major mental health disorders such as bipolar manic episodes ( 20 ) . Another meta-analysis examined the relationship between MTHFR (C677T) genotype and depression diagnosis and reported that those homozygous for the Tallele of the MTHFR gene compared with CC individuals had higher odds of being diagnosed with depression ( 19 ) . These findings are in line with previous studies showing those with the TT genotype of MTHFR genotype have a higher chance of being diagnosed with depression. Taking into consideration MTHFR genotype and folate status may be useful in clinical settings when assessing depressive disorders. Premenstrual depression is a major symptom of PMDD, a diagnostic category of depressive disorders classified in the 5th Diagnostic and Statistical Manual of Mental Disorders ( 4 ) . PMDD is also coded as a gynaecological diagnosis in the WHO’s International Classification of Diseases ( 42 ) . The estimated prevalence of PMDD in the USA is around 8 % ( 43 ) , and in southern Brazil, this estimate goes up to 17·6 % ( 44 ) . These rates continue to rise, worldwide. Outside of the context of menstrual disorders, women who suffer from premenstrual symptoms, are more likely to suffer from postpartum depression as well ( 45 – 47 ) . Interestingly, there seems to be inconsistent evidence for folate as a therapeutic factor for postpartum depression. Some epidemiological studies have found no relationships between folate and postpartum depression ( 48 ) , while systematic reviews have found evidence for folate as a therapeutic measure for perinatal depression ( 49 ) . These inconsistencies may be due to genetic variation in folate metabolism and due to the increased intake of folic acid during the perinatal period and fortification of folic acid in some countries. Despite the mandatory folate fortification of wheat flour and cornmeal in North America, to our knowledge, there is no research directly assessing populations with and without fortification and mental health outcomes. Given the high prevalence of depression as a symptom within menstrual disorders, and in the context of women’s health, it is vital to find minimally invasive lifestyle therapies, such as dietary changes, to decrease the risks of depression in these populations. The present study has several strengths as well as some limitations. We used different measurements of folate including dietary intake and genetic variation in folate metabolism. Using genetic measures of folate metabolism provides an unbiased approach to evaluating the role of folate in premenstrual symptoms. However, the TNH study did not measure plasma or red blood cell folate concentrations, which could have provided a more comprehensive view of folate status. Though participants of the TNH study belonged to one of the three major ethnic groups in Canada, we are left with minimal information on those from Indigenous communities and black individuals. The GHLQ from the TNH study also gathered a comprehensive set of information on each participant’s lifestyle habits, which allowed us to adjust for multiple factors that are recognised to affect both premenstrual symptoms and folate status, minimising the likelihood of residual confounding. Moreover, the association between a polymorphism and nutrient levels is not subject to the limitations of confounding and reverse causation inherent in examining associations between lifestyle factors and health outcomes ( 50 ) . Thus, associations found between MTHFR C677T and premenstrual depression add to the evidence implicating folate and its metabolites in this condition. In conclusion, the current study’s findings indicate that MTHFR genotype may be associated with premenstrual depression specifically in those who have lower intakes of dietary folate. Future research should continue to assess the associations between folate and other B-vitamins and premenstrual depression in intervention settings in women with premenstrual symptoms, PMS and PMDD and account for MTHFR genotype. Furthermore, more research utilising precision nutrition in this field is necessary to provide personalised dietary recommendations.