Of the 1228 adults who attended motivational lectures, 226 subjects consented to participate in the study and were randomly allocated to two groups of equal size. After attrition, the final sample sizes were 87 in both the yoga and control groups. Inclusion criteria were: (a) normal healthy volunteers, (b) age 18–71 years, (c) literacy and (d) scores less than 4/5 in the General Health Questionnaire. Exclusion criteria were: (a) subjects with any ailment, (b) smoking and (c) substance abuse. Source of subjects: Normal adults were recruited from five different locations in Bangalore after public talks at different institutions such as colleges, health clubs, Rotary Clubs, Lion’s clubs and big apartment complexes. Informed consent was obtained from all the subjects who participated in the project and also from the institutional heads where the classes were conducted. The institutional ethical committee of SVYASA cleared the project proposal.

This is a prospective, randomized, single-blind, controlled study aiming to compare the efficacy of yoga (Y) and physical exercise (PE) in normal healthy volunteers in a South Indian population. Introductory lectures were arranged in public centers such as colleges, health clubs, Rotary clubs, Lion’s clubs and apartment complexes. The classes were planned in five different centers in the city of Bangalore. Two hundred and twenty-six persons who consented to participate in the study and satisfied the inclusion and exclusion criteria were randomly allotted to two groups by using five random number tables (different table for each center) generated from the random number generator program.[23] The experimental group was given Y practices and the control group was given PE for one hour daily on empty stomach (6 to 7 a.m.). The classes were conducted six days a week for eight weeks and attendance was maintained by the teachers. Trained experts (in yoga for the Y group and PT for the PE group) conducted parallel sessions for the two groups in different rooms in the same venue. It was ensured that there was no interaction between the subjects. The tests were self-administered before and eight weeks after the intervention. Arrangements were made for the subjects to sit in a quiet place free from distractions and influence from other people.

Masking: The answered questionnaires were coded and kept away for future scoring. A psychologist who was not involved in the subject allocation or supervision of the classes scored the questionnaires which were decoded only after the scoring of both the before and after data was completed.

 

The set of physical exercises were standard execises[28] meant to provide mild to moderate activity designed by experts in physical education.

 

The scoring of the questionnaires was carried out as per the instructions in the manuals. The structure of these questionnaires is described below:
1. VPI evaluates the Sattva, Rajas and Tamas gunas by using a 7-point Likert-type scale. Scores for the gunas are obtained by adding the responses for the items for a guna and then dividing by the number of items for that mode. For each subscale, a higher score indicates a greater predominance of that mode. The minimum and maximum possible scores for the three domains range from 1–7.
2. GHQ: This 28 item test using a binary method of scoring (0, 0, 1, 1) yields an assessment on four robust subscales: somatic symptoms (SS), anxiety and insomnia (AI), social dysfunction (SF) and severe depression (SP). A sum of the scores for these four subscales gives the score for total health. The lower the scores in the GHQ, the better the state of health. The cut-off scores for the GHQ used for this study were 4 or 5 (4/5).[22]

 

Data was analyzed using the SPSS package version 10.0. Based on a previous study,[29] the effect size was calculated to be 0.8. With a power of 0.8 and alpha set to 0.05, the minimum sample size was found to be 164. This calculation was done using G power.[30] The size of the sample actually used was 174.
Data at baseline was assessed for normal distribution using Shapiro-Wilk’s test for both the groups. Independent samples t-test was done for checking homogeneity of baseline scores of the two groups. Paired samples t test and independent samples t test were used for VPI which had normally distributed data and Wilcoxon’s signed ranks and Mann Whitney U tests were used for GHQ data which were not normally distributed. An independent samples t test was done to analyze between the groups and paired samples test within groups. The effect size of the study (mean A – mean B)/ standard deviation (SD) of difference scores) is an absolute measure of the difference that exists between the populations for a parameter, a concept first introduced by the sociologist, J. Cohen.[31] As the study population had a wide age range, statistical analysis was also carried out by grouping them as juniors (age ≤ 24 years) and seniors (age > 24 years) based on the median age. The independent samples t-test for between groups and paired samples t test for within groups were conducted for the two age groups. The data was also analyzed using gender as a factor.