578 participants who had completed at least 97% of the survey indicated that their menstrual cycle had negatively affected their mood or mental health. Three participants were excluded as optional free text responses indicated they were in the menopause or menopause transition. Of the remaining participants, 91.70% (n = 530) had complete data, and data from these participants were analyzed (N = 530). Comparison groups for the current study were created using the PSST, which was composed of 20-items (α in the current sample = 0.93). 32.45% (n = 172) met the threshold for no or mild PMS, 46.79% (n = 248) met the threshold for moderate to severe PMS, and 20.75% (n = 110) met the threshold for PMDD.
Sociodemographic characteristics
Participants’ sociodemographic information can be found in Table 1. The average age was 35.85 (SD = 7.28), with the majority of respondents identifying as female (95.66%, n = 507), and being white (94.15%, n = 499). The majority of participants reported their highest educational attainment as at least an undergraduate degree (71.89%, n = 381) and were employed (either part-time, full-time, or self-employed; 83.02%, n = 440). Most participants had an annual household income, before tax, of between £35,001 and £65,000 (35.85%, n = 190) per annum.
A summary of group differences in sociodemographic characteristics can be found in Multimedia Appendix 1. There was a significant group difference in employment characteristics, with a higher proportion of individuals in full-time employment in the no or mild PMS group than both the moderate to severe and PMDD groups (χ2 = 11.82, df = 2, p = 0.003, φc = 0.15). No other significant group differences were observed.
Table 2
Sociodemographic characteristics
Gender
Female, n (%)
Non-binary, n (%)
Other, n (%)
Prefer not to answer, n (%)
|
507 (95.66)
13 (2.45)
9 (1.70)
1 (0.19)
|
Ethnicity
White, n (%)
Mixed or multiple ethnic groups, n (%)
Asian or Asian British, n (%)
Black, Black British, Caribbean or African, n (%)
|
499 (94.15)
15 (2.83)
14 (2.64)
2 (0.38)
|
Highest educational attainment A
< GCSE or equivalent, n (%)
GCSE or equivalent, n (%)
Advanced level or equivalent, n (%)
Undergraduate degree, n (%)
Postgraduate degree, n (%)
Other, n (%)
Prefer not to say, n (%)
|
5 (0.94)
35 (6.60)
98 (18.49)
181 (34.15)
200 (37.74)
8 (1.51)
3 (0.57)
|
Employment
Employed full-time, n (%)
Employed part-time, n (%)
Self-employed, n (%)
Unemployed, n (%)
Parental leave or taking time off to care for a family member, n (%)
Voluntary work, n (%)
Student, n (%)
Prefer not to answer, n (%)
|
239 (45.09)
146 (27.55)
55 (10.38)
39 (7.36)
34 (6.42)
12 (2.26)
61 (11.51)
6 (1.13)
|
Household income
Less than £15,000, n (%)
£15,001-£25,000, n (%)
£25,001-£35,000, n (%)
£35,001-£45,000, n (%)
£45,001-£55,000, n (%)
£55,001-£65,000, n (%)
£65,001-£75,000, n (%)
£75,001-£85,000, n (%)
Over £85,000, n (%)
Prefer not to say, n (%)
|
39 (7.36)
43 (8.11)
65 (12.26)
72 (13.58)
53 (10.00)
65 (12.26)
37 (6.98)
35 (6.60)
71 (13.40)
50 (9.43)
|
Key. GCSE: General Certificate of Secondary Education; A: GCSE and Advanced level are academic qualifications taken by secondary education students in the United Kingdom, in the 11th and 13th year of education, respectively. Note: Percentage total for employment exceeds 100% as participants could select multiple answer options |
Mental health characteristics
The mean WEMWBS score (α in the current sample = 0.93) across the whole sample was 40.95 (SD = 8.72). A one-way ANOVA revealed a significant differences in WEMWBS scores between all the groups (F(2, 527) = 22.70, p < .001, η2 = 0.08), with lowest wellbeing scores in the PMDD group (mean = 36.62, SD = 9.65) relative to both other groups. Further, significantly lower wellbeing scores were observed in the moderate to severe PMS group (mean = 41.09, SD = 7.45) than in the no or mild PMS group (mean = 43.51, SD = 8.71).
The entire sample endorsed at least one premenstrual symptom to at least a mild severity. The most frequently endorsed premenstrual symptoms screened for using the PSST in the overall sample were physical symptoms (eg, breast tenderness, headaches; 96.42%; n = 511; Fig. 1) followed by psychological symptoms related to increased anger and irritability (95.85%, n = 508). The symptom most frequently reported as severe in the overall sample was fatigue/lack of energy (36.23%, n = 192). 97.17% (n = 515) of the overall sample stated that premenstrual symptoms had affected their functioning to at least a mild extent in at least one domain. The most frequent domain of functional impairment was in work/studies (83.40%, n = 442), with the most severely impaired domain reported as romantic or intimate relationships (15.85%, n = 84).
In the no or mild PMS group (Fig. 2), physical symptoms (eg, breast tenderness, headaches) were the most frequently endorsed symptoms (93.02%, n = 160), with fatigue/lack of energy being most frequently endorsed as severe (11.05%, n = 19). The most frequent domain of functional impairment arising from premenstrual symptoms was work/studies (68.02%, n = 117). Social life was also the domain which was most frequently endorsed as being severely impacted (1.16%, n = 2).
In the moderate to severe PMS group (Fig. 3), the most frequently reported symptoms were anger irritability and tearfulness/sensitivity to rejection (99.19%, n = 246). The symptom most frequently endorsed as severe was fatigue/lack of energy (40.73%. n = 231). The most frequently impaired domain of functioning reported was home responsibilities (95.15%, n = 205), with the domain most endorsed as severe being romantic or intimate relationships (7.66%, n = 19).
In the PMDD group (Fig. 4), the most frequent symptom, endorsed by all individuals in this group (100%, n = 110), was anxiety/tension. The most frequently endorsed severe symptoms were depressed mood/hopelessness and tearfulness/increased sensitivity to rejection (70.91%, n = 78). The domain of functioning which was reported to be most frequently impaired was home responsibilities (98.18%, n = 108), and the most severely impaired domain of functioning being romantic or intimate relationships (58.18%, n = 64).
Significant group differences were observed in the frequency (Hs(2) ≥ 8.47, Ps≤0.014, η2s ≥ 0.02) and severity (Hs(2) ≥ 22.91, Ps<0.001, η2s ≥ 0.04) of premenstrual symptoms and associated functional impairment (Multimedia Appendix 1).
Significant differences were observed for the frequency of all symptoms between the no or mild PMS and the moderate to severe PMS groups, with the latter having a higher proportion of individuals who endorsed each PMS symptom (Us≥12622.00, Ps≤0.11, rs≥-0.12). Similarly, when comparing the frequency of PMS symptoms between the no or mild PMS and PMDD groups, the PMDD group had a higher proportion of individuals who endorsed each symptom (Us≥19950.00, Ps≤0.019, rs≥0.14), aside from physical symptoms, where no significant difference was observed (U = 8972.00, P = .052, r=-0.12). Significant differences were identified between the moderate to severe PMS and PMDD groups for the frequency of the following PMS symptoms: decreased interest in work (U = 12568.00, P = .008, r=-0.14), home (U = 12293.00, P = .002, r=-0.16) and social activities (U = 12279.00, P = .001, r=-0.17), and in difficulty concentrating (U = 12733.00, P = .018, r=-0.12).
With regard to differences in the frequency of severely reported symptoms between groups, both the moderate to severe PMS and PMDD groups more frequently reported severe symptoms than the no or mild PMS group for all symptoms (Us≥3027.00, Ps≤.010, rs≥-0.14). Moreover, there were significant differences between the moderate to severe PMS and PMDD groups, with all symptoms more likely to have been reported as severe in the PMDD group than the moderate to severe PMS group (Us≥6168.00, Ps≤.034, rs≥0.11).
Regarding the frequency of functional impairment, there were significant differences between the no or mild PMS group and both the moderate to severe PMS and PMDD groups, with the moderate to severe PMS and PMDD groups being more likely to endorse impairments in every domain of functioning relative to the no or mild PMS group (Us≥5072.00, Ps<.001, rs≥-0.12). Between the moderate to severe PMS and PMDD groups, there were significant differences in the frequency of impairment in relationships with co-workers (U = 11873.00, P = .017, r=-0.13) and home responsibilities (U = 12953.00, P = .050, r=-0.10), with the PMDD group being more likely to have experienced impairment in these domains.
In terms of group differences in severe functional impairment, the moderate to severe PMS group had a higher proportion of individuals reporting severe impairment in the domains of family (U = 19818.00, P < .001, r=-0.25) and home responsibilities (U = 20592.00, P = 0.30, r=-0.48) relative to the no or mild PMS group. Furthermore, the PMDD group reported severe functional impairment in all domains more frequently than both the no or mild and the moderate to severe PMS groups (Us≥4011.00, Ps<.001, rs≥-0.29).
Help-seeking characteristics
Data was collected on formal help seeking (ie, through a health care provider such as a National Health Service (NHS) GP) and online self-help seeking. In the overall sample, 64.91% (n = 344) had sought help (either only-formal, only-online, or blended online and formal; Fig. 5). Overall, 35.09% (n = 186) of participants stated that they had visited a HCP to discuss their mental health symptoms in relation to their menstrual cycle. Furthermore, in the overall sample, 57.55% (n = 305) had searched online for help. In the overall sample, 7.36% (n = 39) had only sought formal help, 29.81% (n = 158) had only sought online help, and 27.74% (n = 147) had sought both formal and online help.
Within the groups, 47.67% (n = 82) of the no or mild PMS group, 66.94% (n = 166) of the moderate to severe PMS, and 87.27% (n = 96) of the PMDD group had sought help. A chi-squared test explored group differences in help-seeking behaviours, revealing significant group differences (χ2 = 79.21, df = 6, p < 0.001, φc = 0.39) in not seeking help and blended help-seeking. The proportion of those engaged in blended help-seeking increased with symptom severity, with the PMDD group having the highest proportion of blended help-seekers (55.45%, n = 61) compared to both the no or mild (25.81%, n = 64) and moderate to severe PMS (12.79%, n = 22) groups. Similarly, the highest proportion of those who had sought no help were the no or mild PMS group where symptom severity is lowest (52.33%, n = 90) as compared to the moderate to severe (33.06%, n = 82) and PMDD groups (12.73%, n = 14). There were no significant group differences between groups in the proportion of individuals who had only sought online or only sought formal help.
The most frequent type of help searched-for online was looking up mental health symptoms related to the menstrual cycle (85.57%, n = 261), followed by looking up information related to available treatment options (including therapy, counselling, and online cognitive behavioural therapy; 39.67%, n = 121), taking an online mental health test (34.43%, n = 105), and searching for relevant apps (33.44%, n = 102).
Group differences in the type of help sought online (n = 304) were explored using chi-squared tests. There were significant group differences in online searching for treatment options (χ2 = 9.79, df = 2, p = .007, φc = .18), looking online for a mental health assessment (χ2 = 7.92, df = 2, p = .019, φc = .16), looking online for peer support (χ2 = 12.79, df = 2, p = .002, φc = .21), and looking online for relevant apps (χ2 = 27.22, df = 2, p < .001, φc = .30). Specifically, a significantly higher proportion of the PMDD group had searched online for information on treatment options, an online mental health test, peer support, and relevant apps than both the no or mild and moderate to severe PMS groups.
Overall, with regard to the quality of consultations with HCPs (n = 186), only 21.51% (n = 40) of participants stated that they felt their mental health symptoms related to the menstrual cycle were taken at least moderately seriously. 45.16% (n = 84) of participants who had visited a HCP stated that their mental health symptoms related to their menstrual cycle were not taken seriously at all, and a further 33.33% (n = 62) stated that they were only taken slightly seriously. No significant group differences were found for perceived quality of consultations with HCPs.