Prediction of Mental Illness using Indian Astrology: Cross-sectional Findings from a Prospective Study

Abstract

Background: Studies involving astrology and psychiatry have mostly found conflicting results, with astrology being criticized as unscientific and also, lack of an objective assessment method for scientifically testing astrology. The current study attempts to test the predictive ability of astrology using the Indian system for the first time in identifying mental illness.

Methodology: The sample comprised of 150 subjects, with 75 having mental illness and 75 without mental illness. Keeping the subjects and astrologers blind to each other, a computer generated birth chart data using the subject’s gender, date, time and place of birth was given for interpretation. Predictions were matched with the clinical condition at intake.

Results: Kappa coefficient was used which found a moderate agreement in the prediction of life time mental illness (k=0.560, p=.001) and a substantial agreement in predicting current state of mental illness (k=0.626, p=.001). Viewed as a diagnostic test, astrology showed, a fairly good sensitivity (77.3%), specificity (78.7%), positive predictive value (78.4%) and negative predictive value (77.6%) in identifying the life-time and current presence or absence of mental illness. However, the study showed poor match while predicting symptom cluster (34.6% in mental illness, 42.6% in normal subjects) and onset of symptoms (28%).

Conclusions: Overall, the evidence seems to point towards astrology, as practiced in this study, modestly predicting presence of mental illness. Caveats concerning the methodology, statistics including poor inter-astrologer agreement makes it difficult in reconciling conflicting findings. Hence the findings in this study are at best tentative and need more extensive enquiry.