The sample size was calculated from an earlier study that compared the effect of two different drugs in menopausal women (as there are no studies on yoga). Using the pre-post mean and SD values from the vasomotor outcome variables from that study, an effect size of 0.52 was calculated.32 Using this value of effect size and the values for 'alpha* and power at 0.05 and 0.8, respectively, a sample size of 108 was derived. Of a total of 201 women experiencing menopausal symp­toms screened, 120 women (married or single) who satisfied the inclusion criteria of (a) age between 40 and 55 years and (b) serum FSH level equal to or higher than 15 miu/ml on the sixth day of the menstrual cycle if she was menstruating regularly or at the time of recruitment if she had stopped menstruating or had irregular cycles were selected for the study. Women who had undergone hysterectomy with retained ovaries were also included. Exclusion criteria were (a) women who were practicing yoga for a month or more, (b) women with no knowledge of English, (c) women with less than high school education, (d) women taking HRT, (e) women who underwent any surgery in past 3 months, (f) those with gynaecological problems like endometriosis, fibroids, ovarian cysts, prolapsed uterus etc., (g) women with other medical disor­ders (like hypertension, diabetes mellitus, hypo/hyperthyroidism) and (h) those on psychiatric medication.

The study was conducted at the Yoga University, Swami Vivekananda Yoga Research Foundation (SVYASA) in Bangalore city. The subjects were recruited through advertisements and giving talks about the benefit of these practices in women's organisations, clubs and organisations such as lioness clubs. They were also contacted through banners, newspaper adver­tisements and circulation of pamphlets apart from references through word of mouth. Some women were recruited through gynaecological clinics as they consulted the doctor for their menopausal symptoms. In all, they were recruited from 14 different areas of Bangalore and classes were conducted at 14 nodal centres of SVYASA in different parts of the city.

The institutional review board and ethical committee of the University Swami Vivekananda Yoga Research Foundation, Bangalore, sanctioned formal approval. The research staff answered queries and the participants then made an informed independent decision about participating in the study.

This was a prospective, randomised controlled trial (RCT) wherein 120 participants were randomly divided into two groups: one group practised integrated approach of yoga ther­apy (IAYT) and the other practised a set of physical exercises. The women who satisfied the inclusion criteria were registered in different nodal centres by using pre-labelled envelopes to avoid selection bias; roll numbers were allotted and these numbers were randomly divided into two groups using a computer-generated random number table (hltp://www. randomizer.org) prepared for the specific number of participants available in the centre. Participants were assessed for the cognitive tests before and after 8 weeks of intervention. Both the yoga and control groups were given their respective set of practices for 1 hour of intervention per day, 5 days per week for 8 weeks, by specially trained instructors for both yoga and non-yogic physical exercise groups,

As this is an interventional study, this could not be a double-blind study, but attempts were made to blind and mask wher­ever feasible to reduce the bias. The statistician who did the randomisation of the serial numbers of participants and the final analysis was blind to the source of the data. The answer sheets for the six-letter cancellation test (SLCT) and Punit Govil intelligence Memory Scale (PG1MS) were coded and kept away for final analysis and were decoded only after complete analysis. The memory tests were administered by a psychologist (who was not involved in interacting with the participants) to the whole group before randomisation. Care was taken to arrange the timings and venue of the classes for the two groups suitably to avoid interaction and exchange of information and techniques among the participants of the two groups.

A. checklist of three of the major symptoms of climacteric (vasomotor symptoms checklist [VCL] 1, hot flushes; VCL 2, night sweats and VCL 3, sleep disturbances), with severity scoring ranging from 0 to 3 (0 being absence of that symptom and 3 being severely suffering from that symptom), was used to assess the vasomotor symptoms for all participants.

Six-letter cancellation test (SLCT) for adults is a paper-and-pencil test that uses a letter cancellation task that measures cognitive functions such as selective and focused attention, visual scanning as well as activation and inhibition of rapid responses. It consists of a test worksheet that specifies six target letters to be cancelled and has a "working section*, which consists of letters of the alphabet, arranged randomly in 22 rows and 14 columns. The participants are asked to cancel as many of the six target letters as possible in a specified time of 90 seconds.34 The total number of cancellations and wrong cancellations are scored, and the net scores are cal­culated by deducting wrong cancellations from the total at­tempt. This test has been evaluated for its reliability and validity based on standard criteria. Reliability has been ascer­tained based on (a) temporal stability and (b) internal con­sistency.33 The content validity of this test is adequate for the purpose for which it is intended.56 The normal value for healthy Indian adults for SLCT is 38 ± 6.M

PGIMS is a battery of ten memory tests, which measures the remote memory, recent memory, mental balance, two tests for attention and concentration (for digit memory and reverse digit memory), delayed recall, immediate recall, retention for similar pairs, retention for dissimilar pairs, visual retention and recognition test. The participant is supposed to write the responses to the questions asked by the administrator. Of the ten tests, eight tests are verbal, one test pertains to geometrical drawing and one on recognising objects. The reliability of this scale has been tested, and the norms for adults (>20 years) with no psychiatric/neurological illnesses are available. PGIMS is incorporated as one of the important tests to evaluate cog­nitive functions and organic brain dysfunctions. Administration takes 15-20 minutes per participant The test retest reliability on 40 subjects ranged between 0.70 and 0.84 for organic psychotic groups, 0.48 and 0.84 for neurotic group. On the whole, however, an increase of four points was observed on repeated testing. For these two groups, split-half reliability was found to be 0.91 and 0.83, respectively.37

The control group practiced a set of exercise program comprising easy (nonsweating) body movements supervised by physical trainers for the same duration of 1 hour daily, 5 days a week for 8 weeks. The exercise involved the stretching of the arms, legs and spinal twists, strengthening the muscles around knee joints, shoulder joints, neck and wrist joints. Lectures and individual counseling was given on conventional modern medical concepts about healthy lifestyle including diet, exercise and physiology of menopause and stress management techniques. Details are given in Table 1.

Data of 108 (leaving the 12 dropouts of 120} participants were scored as per the instructions in the manual by a psychologist and were analysed by the statistician using SPSS version 10.0 (SPSS Inc., Chicago, IL, USA). The test of normality was carried out using Kolmogorov-Smirnov test as the sample size was above 50. Because the data were not normally distributed, nonparametric tests were used: Mann-Whitney test for between group comparison and Wilcoxon signed-ranks test for within group comparison. The baseline values for all the variables in both the groups were compared using Mann—Whitney test Effect sizes were calculated to measure Ae magnitude of difference of parameters between the two groups.42 Correlations were checked by Spearman's correlation coefficient, as the data were non-normal For the design of this study (one within-subjects and one between-subjects factor), there were no ancillary analyses.