In a recently published study, Dr Journal of Affective DisordersA group of researchers measured the point prevalence of premenstrual dysphoric disorder (PMDD) using both definite and provisional diagnoses.
PMDD, a significant mental health concern, has evolved from a more studied condition in previous Diagnostic and Statistical Manual of Mental Disorders (DSM) editions to a recognized disorder in DSM-5 and International Classification of Diseases (ICD)-11. This disorder, characterized by severe emotional and physical symptoms associated with the menstrual cycle, is now understood to severely impact quality of life and is associated with an increase in suicide. Diagnosis depends on the specific timing and severity of symptoms, DSM-5 and ICD-11 emphasize luteal phase symptoms including mental changes and physical discomfort.
The diagnosis process, ideally based on potential symptom ratings over two menstrual cycles, varies between definite and tentative depending on symptom reporting methods. More research is needed to better understand PMDD’s etiology, treatment efficacy, and long-term effects on mental and physical health.
About the study
The researchers conducted a thorough search for observational studies that examined the prevalence of PMDD in women from menopause to menopause, adhering to DSM or ICD diagnostic criteria. Studies relying solely on self-diagnosis were excluded due to their unreliability. They excluded studies focusing on participants from health services such as gynecological clinics to avoid inflated prevalence figures. When encountering multiple studies from the same sample, the one with the largest sample size was selected. There were no language restrictions, and Google Translate was employed for non-English reports.
The search spanned several databases – Embase, PsycINFO, MEDLINE, and PubMed – from their inception to March 2021. Two authors independently screened study abstracts, selecting relevant full-texts for detailed examination A variety of terms related to PMDD were included in the search to ensure a comprehensive literature review.
For data analysis, two authors discussed important information such as study setting, sample size, and assessment methods, resolving any disagreements by discussion with a third author. Each study was assessed for bias using a specific risk of bias tool.
Statistical analyzes were performed using R statistical software, employing a random-effects model to account for variation across studies. They evaluated the variance using i2 explored its potential sources with statistics and subgroup analysis and meta-regression. Studies were further classified based on the method of diagnosis (provisional or confirmed) and the use of questionnaires.
Results of the study
In the current comprehensive search, the researchers identified 12,340 records, eventually narrowing down to 44 studies that met their criteria. This study represented 48 independent samples with a total of 50,659 participants, ranging from 62 to 8,694 per sample and a mean sample size of 1151. Participants’ ages varied between 14.3 and 38.6 years. These studies were globally distributed, with significant numbers from Asia, Europe, and North America. Most studies used DSM criteria for diagnosis, with no studies using ICD-11 criteria.
The researchers noted that only a minority of studies employed prospective ratings for a definitive diagnosis of PMDD. There was a significant difference in mean sample size between studies with confirmed (798) and provisional (1255) diagnoses. Importantly, one study reported no cases of PMDD in its sample. The researchers found that the pooled prevalence of PMDD was 3.2% for confirmed cases and 7.7% for provisional diagnoses, although there was significant variation (I).2 = 99%).
Risk of bias varied between studies with scores ranging from 3 to 9. Notably, all studies were consistent in their data collection methods and obtained data directly from participants. A high prevalence was observed in studies with provisional diagnoses, African samples, and university settings. Also, a negative correlation between prevalence rate and risk of bias was observed.
The use of questionnaires in diagnosing PMDD also influenced prevalence rates. Studies using validated questionnaires reported lower prevalence than studies that did not use a standardized questionnaire. Of the studies that reported provisional diagnosis, 22 used a validated questionnaire.
A particular study by de la Gandara Martín and de Diego Herrero in 1996, which reported an unusually high prevalence of 30.5%, was considered an outlier and excluded from further analysis. This adjustment reduced the pooled prevalence by 2.7% and reduced the variance.
Focusing only on community-based samples that strictly adhered to DSM criteria, including those tracking two cycles of symptoms, resulted in more refined analyses. This restriction revealed a pooled prevalence of 1.6% with significantly less heterogeneity for six samples (I2 = 26%). This finding underscores the importance of strict adherence to diagnostic criteria and sample selection in determining the prevalence of PMDD.