 |
 |
|
|
|
The diagnostic criteria for premenstrual dysphoric disorder (PMDD), as defined in the DSM-IV, are much stricter than those for premenstrual syndrome (PMS) (1). To apply the DSM-IV criteria for PMDD, women must prospectively chart symptoms daily for at least 2 consecutive symptomatic cycles, and their chief complaints must include 1 of the 4 core mood symptoms (that is, irritability, tension, dysphoria, and lability of mood) and at least 5 of the 11 total symptoms. The charting of symptoms should clearly demonstrate premenstrual worsening and remission within a few days after the onset of menstruation (“on-offness”) (2). A change in symptoms from the follicular to the luteal phase of at least 50% is suggested for the diagnosis of PMDD (3,4). Results from a study of women with PMS who sought medical attention suggest that the DSM-IV criteria for PMDD may be too strict (5). Subjects failing to meet the criteria may still have substantial premenstrual worsening of symptoms. Prospective daily self-rating of symptoms, using reliable and valid instruments, is essential in making the diagnosis. To date, there is still no consensus among investigators as to the best instrument for confirming prospectively the diagnosis of PMDD, nor is there consensus as to the most appropriate instrument to measure treatment effects in clinical trials (6). Previous studies have shown that using a single-item visual analog scale (VAS) for each of the 4 core premenstrual mood symptoms is a reliable and valid method of prospective data collection (3,4). Most of these studies, though, predate the introduction of the DSM-IV diagnostic criteria for PMDD, and previous VASs are not aligned with these criteria. We further report here on the reliability and validity of the VASs that were revised to better reflect the DSM-IV definition of PMDD. MethodsThe analysis used concurrent data from women’s scores on the revised VASs and the validated Premenstrual Tension Syndrome Observer (PMTS-O) rating scale collected in 4 randomized controlled trials (n = 1208) evaluating the efficacy of different treatment options with paroxetine. In all these trials, women with premenstrual symptoms that fulfilled the DSM-IV diagnostic criteria for PMDD signed a written informed consent form that was approved by the institutional review board at participating centres. The validated PMTS-O scale allows an observer to assess symptoms in 10 domains: irritability or hostility, tension, efficiency, dysphoria or moodiness, motor coordination, mental–cognitive functioning, eating habits, sexual drive and activity, physical symptoms, and social impairment (7). VASs, together with the PMTS-O scale, have been shown to be a valid, reliable, and sensitive measure for the severity of premenstrual symptoms (3). The PMTS-O scale has also been used in subjects with premenstrual symptoms to establish symptom severity for inclusion in clinical studies (8–11) and to evaluate treatment response in clinical trials (12–15). In 3 trials with regular visits scheduled during the follicular phases (that is, each within 3 days after onset of menses), 1030 participants were randomized to treatment groups of continuous treatment with paroxetine (12.5 mg daily), paroxetine (25 mg daily), or placebo (16). The trials were conducted according to an identical protocol with minor language adaptations for different geographic regions of Europe and North America. Participants completed a self-rating set of 11 VASs daily throughout 6 menstrual cycles (2 screening, 1 baseline, and 3 treatment cycles). Single-item VASs were used to measure each of the 4 core mood symptoms (depressed mood, tension, affective lability, and irritability), as well as the 7 additional clusters of symptoms (decreased interest in usual activities, difficulty with concentration, lack of energy, change in appetite, change in sleep pattern, feeling out of control, and physical symptoms), in accordance with the DSM-IV criteria for PMDD (1). During regular visits, independent observers also assessed participants by using a retrospective data collection approach with the PMTS-O scale, prompting for the severity of premenstrual symptoms associated with the late luteal phase of the preceding menstrual cycles. In one trial with regular visits scheduled during the luteal phases, 178 participants were randomized to treatment groups receiving continuous treatment with paroxetine (20 mg daily), intermittent treatment with paroxetine (20 mg daily during the luteal phase only), or placebo (17). Participants completed a self-rating, slightly modified (that is, a Swedish version) set of 10 VASs daily for 6 cycles. Single-item VASs were used to measure each of the 4 core mood symptoms (depressed mood, tension, affective lability, and irritability) as well as 6 additional clusters of symptoms (mood swings, bloatedness, breast tenderness, lack of energy, food cravings, and menstrual pain). A single observer used a prospective data collection approach with the PMTS-O scale to assess participants during their luteal phases. In all trials, each VAS consisted of a 100-mm horizontal line with vertical line anchors at each end. The anchors were 0 = “not at all” (that is, “the way you normally feel when you don’t have premenstrual symptoms”) and 100 = “extreme symptoms” (that is, “the way you feel when your premenstrual symptoms are at their worst”). Data Analysis All analyses were based on intention-to-treat. For the analyses, data from the 3 trials with follicular phase visits were pooled on the basis of their identical protocols and the homogeneous results from the reliability analyses of individual trial data. Data from the trial with luteal phase visits were analyzed separately. The VAS scores on the 4 core symptoms were averaged to create a mean score to represent mood symptoms (that is, VAS mood score) (4). We calculated the VAS total scores (average of 11 symptom scores) only for the trials with follicular phase visits. For the trial with luteal phase visits, the VAS total scores (average of 10 symptom scores) were not derived because they were not comparable with those derived from the other trials. We calculated a late luteal phase score (that is, VAS mood or VAS total), using the average daily VAS scores of the 5 days prior to menses, and a follicular phase score, using the average of daily VAS scores from postmenses days 6 to 10. The PMTS-O scale scores 10 domains (outlined above) with severity ranging from 0 to 4 in 8 subscales and 0 to 2 in 2 domains, for a maximum score of 36. A PMTS-O mood subscore includes items from question 1 (irritability or hostility), question 2 (tension), and question 4 (dysphoria or moodiness) and ranges from 0 to 12. Change scores from both VAS and PMTS-O scales were used to assess sensitivity to symptom worsening or improvement. A change score was derived by subtracting a baseline score from the corresponding score at study end (that is, the third treatment cycle in all trials). For participants who dropped out, we used the score of the last cycle before early termination. Cronbach’s alpha coefficient was used at baseline to evaluate the internal consistency of the VAS mood items. Baseline VAS scores were summarized for the late luteal and follicular phases across treatment options. To evaluate construct validity, we compared VAS mood scores for subgroups of participants with varying premenstrual symptom severity according to a priori defined PMTS-O thresholds. A PMTS-O total score greater than 27 has been suggested for severe symptoms, between 18 and 27 for moderate symptoms, and between 10 and 17 for mild symptoms (18,19). Analysis of variance was used to compare subgroups. Pearson’s correlation was used to compare the DSM-IV VAS and PMTS-O scores. Both absolute scores at baseline and change scores were compared. Individual items, mood domain scores, and total scores were compared. To examine the construct validity of the VAS items relative to those of the PMTS-O scale, pairwise correlation coefficients were examined between the 4 core mood symptoms in the VAS scale and the corresponding 3 symptoms from the PMTS-O scale. Correlation coefficients from 0.3 to 0.5 indicated moderate association; from 0.5 to 0.7, strong association; and above 0.7, excellent association (20). ResultsData from 1208 participants were used in the analyses. In the trials with follicular phase visits, 1030 participants contributed data to the baseline analyses and 934 (88%) to the change score analyses. For the trial with luteal phase visits, the corresponding figures were 178 and 165 (93%), respectively (Tables 1 and 2).
The internal consistency coefficient for the luteal VAS mood score was 0.90 with data from the 3 trials with follicular phase visits (ranging from 0.90 to 0.91 across trials) and 0.88 with data from the trial with luteal phase visits. The corresponding coefficients for the follicular VAS mood score were 0.94 (ranging from 0.92 to 0.96 across trials) and 0.91, respectively. These results indicate a high level of internal consistency among the 4 core mood symptoms. The VAS mood scores captured the “on-offness” of the conditions, with the mean VAS mood score ranging from 47 to 57 across treatment options for the late luteal phase and from 5 to 9 for the subsequent follicular phase. Similar results were observed with the VAS total scores (Table 1). According to the construct validity criterion, 15% (n = 153) of participants in the trials with follicular phase visits experienced mild premenstrual symptoms; 68% (n = 682), moderate symptoms; and 17% (n = 171), severe symptoms. The mean VAS mood score was 41 (95%CI, 38 to 44) for participants with mild symptoms, 54 (95%CI, 52 to 55) for those with moderate symptoms, and 70 (95%CI, 66 to 73) for those with severe symptoms. The mean VAS mood score was significantly different across symptom severities (F2,1003 = 75.7, P < 0.0001). Luteal phase VAS scores and corresponding PMTS-O scores were moderately correlated at baseline (P < 0.01) (Table 2). The correlation coefficient between VAS mood and PMTS-O mood scores was 0.48 for the 3 trials with follicular phase visits (ranging from 0.42 to 0.50 across individual trials) and 0.46 for the trial with luteal phase visits. Luteal phase VAS change scores and corresponding PMTS-O change scores were strongly correlated (P < 0.01), indicating similar sensitivity to premenstrual symptom change by both scales. Whether observers assessed symptom severity retrospectively or prospectively, the correlations between the patient-reported VAS scores and the PMTS-O scores were comparable (Table 2). Corresponding items from the 2 scales (for example, VAS depressed mood and PMTS-O dysphoria or moodiness) always attained relatively higher correlation values when compared with correlation between noncorresponding items (for example, VAS depressed mood and PMTS-O tension), suggesting further evidence of the construct validity of VAS items. DiscussionData collected from 4 treatment trials of women with severe premenstrual symptoms indicate that using the revised VASs (which better reflect the current DSM-IV definition of PMDD) provides a reliable measure of premenstrual symptoms when evaluated against the well-validated PMTS-O scale. The results also suggest similar concurrent validity between the daily rating of symptom severity using the patient-reported revised VAS and the observers’ rating scale (PMTS-O), regardless of whether the observers assessed the premenstrual symptoms with a prospective or retrospective data collection approach. The DSM-IV requires women to prospectively chart symptoms daily for a minimum of 2 symptomatic cycles to qualify for the diagnosis of PMDD and hence meet inclusion criteria in clinical trials. It has recently been suggested that this may be impractical as a diagnostic tool in a busy primary care practice. Not only has it been shown that the requirement for prospective charting may act as a deterrent for seeking help, it may also indicate that patients who participate in clinical trials are very different from typical PMDD patients seen in primary care (21). The revised VAS better reflects the current DSM-IV diagnostic criteria for PMDD, and it is also more user-friendly. As such, it can potentially increase the accurate identification of sufferers and also improve their compliance with treatment. Most women with severe PMS or PMDD are likely to seek treatment from their obstetrician-gynecologist or other primary health care provider. At present, there is still no consensus on diagnostic criteria for severe PMS and PMDD (22). Recent attempts at circumventing the need for a prospective daily paper and pencil charting are promising (23,24); however, further validation studies are needed. The data presented here seem to also indicate that, in women with clear-cut severe PMS and (or) PMDD, retrospective assessment may be a clinically acceptable alternative to strict DSM-IV research diagnostic criteria. Funding and SupportFunding for the project was provided by GlaxoSmithKline. AcknowledgementsThe authors thank Brian Hunter, Timothy Rolfe, and Reid Robson for their input and support during the project; Tina Haller, Theresa Chua, and Jenny Huang for their assistance with the data analysis; and Carol Ballantyne and Cindy Tasch for their expert help in preparing the manuscript. References1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed. Washington (DC): American Psychiatric Association; 1994. p 715–8. 2. Endicott J, Amsterdam J, Eriksson E, Frank E, Freeman E, Hirschfeld R, and others. Is premenstrual dysphoria disorder a distinct clinical entity? J Womens Health Gend Based Med 1999;8:663–79. 3. Steiner M, Streiner DL, Steinberg S, Stewart D, Carter D, Berger C, and others. The measurement of premenstrual mood symptoms. J Affect Disord 1999;53:269–73. 4. Steiner M, Steinberg S, Stewart D, Carter D, Berger C, Reid C, and others. Fluoxetine in the treatment of premenstrual dysphoria. N Engl J Med 1995;332:1529–34. 5. Kraemer CR, Kraemer RR. Premenstrual syndrome: diagnosis and treatment experiences. J Womens Health 1998;7:893–907. 6. Born L, Palova E, Steiner M. Premenstrual syndromes: guidelines for treatment. In: Gaszner P, Halbreich U, editors. Women’s mental health. Budapest (Hungary): Section of Interdisciplinary Collaboration, World Psychiatric Association; 2002. p 56–72. 7. Steiner M, Haskett RF, Carroll BJ. Premenstrual tension syndrome: the development of research diagnostic criteria and new rating scales. Acta Psychiatr Scand 1980;62:177–90. 8. Taskin O, Gokdeniz R, Yalcinoglu A, Buhur A, Burak F, Atmaca R, and others. Placebo-controlled cross-over study of effects of tibolone on premenstrual symptoms and peripheral beta-endorphin concentrations in premenstrual syndrome. Hum Reprod 1998;13:2402–5. 9. Brown CS, Ling FW, Andersen RN, Farmer RG, Arheart KL. Efficacy of depot leuprolide in premenstrual syndrome: effect of symptom severity and type in a controlled trial. Obstet Gynecol 1994;84:779–86. 10. Rausch JL, Janowsky DS, Golshan S, Kuhn K, Risch SC. Atenolol treatment of late luteal phase dysphoric disorder. J Affect Disord 1988;15:141–7. 11. Maddocks S, Hahn P, Moller F, Reid RL. A double-blind placebo-controlled trial of progesterone vaginal suppositories in the treatment of premenstrual syndrome. Am J Obstet Gynecol 1986;154:573–81. 12. Condon JT. Investigation of the reliability and factor structure of a questionnaire for assessment of the premenstrual syndrome. J Psychosom Res 1993;37:543–51. 13. Hahn PM, Van Vugt DA, Reid RL. A randomized, placebo-controlled, crossover trial of danazol for the treatment of premenstrual syndrome. Psychoneuro- endrocrinology 1995;20:193–209. 14. Schmidt PJ, Grover GN, Rubinow DR. Alprazolam in the treatment of premenstrual syndrome: a double-blind, placebo-controlled trial. Arch Gen Psychiatry 1993;50:467–73. 15. Su TP, Schmidt PJ, Danaceau MA, Tobin MB, Rosenstein DL, Murphy DL, and others. Fluoxetine in the treatment of premenstrual dysphoria. Neuropsychopharmacology 1997;16:346–56. 16. Yonkers KA, Hunter BN, Bellew KM, Rolfe TE, Steiner M, Heller VL. Paroxetine controlled release is effective in treating premenstrual dysphoric disorder: a pooled analysis of three trials. Obstet Gynecol 2003;101:110S–111S. 17. Landen M, Sorvik K, Ysander C, Allgulander C, Nissbrandt H, Gezelius B, and others. A placebo-controlled trial exploring the efficacy of paroxetine for the treatment of premenstrual dysphoria [poster presentation, 2002]. Washington (DC): American Psychiatric Association; 2002. 18. Smith S, Rinehart JS, Ruddock VE, Schiff I. Treatment of premenstrual syndrome with alprazolam: results of a double-blind, placebo-controlled, randomized crossover clinical trial. Obstet Gynecol 1987;70:37–43. 19. Miner C, Brown E, McCray S, Gonzales J, Wohlreich M. Weekly luteal-phase dosing with enteric-coated fluoxetine 90 mg in premenstrual dysphoric disorder: a randomized, double-blind, placebo-controlled clinical trial. Clin Ther 2002;24:417–33. 20. Streiner DL, Norman GR. Health measurement scales: a practical guide to their development and use, 3rd ed. Oxford (UK): Oxford University Press; 2003. 21. Yonkers KA, Pearlstein T, Rosenheck RA. Premenstrual disorders: bridging research and clinical reality. Arch Women Ment Health 2003;6:287–92. 22. Freeman EW. Premenstrual syndrome and premenstrual dysphoric disorder: definitions and diagnosis. Psychoneuroendocrinology 2003;28(Suppl 3):25–37. 23. Wyatt KM, Dimmock PW, Hayes-Gill B, Crowe J, O’Brien PM. Menstrual symptometrics: a simple computer-aided method to quantify menstrual cycle disorders. Fertil Steril 2002;78:96–101. 24. Steiner M, Macdougall M, Brown E. The premenstrual symptoms screening tool (PSST) for clinicians. Arch Women Ment Health 2003;6:203–9. Author(s)Manuscript received April 2004, revised, and accepted July 2004. 1. Professor, Department of Psychiatry and Behavioural Neurosciences and Department of Obstetrics and Gynecology, McMaster University, Hamilton, Ontario. 2. Assistant Vice President, Research and Director, Baycrest Centre for Geriatric Care; Professor, Department of Psychiatry, University of Toronto, Toronto, Ontario. 3. Statistician, BioMedical Data Sciences, GlaxoSmithKline, Oakville, Ontario. Address for correspondence: Dr M Steiner, Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, ON L8N 4A6 e-mail: mst@mcmaster.ca
1 | 2
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||