Seasonal variation in anorexia nervosa? Some preliminary findings from a neglected area of research


  • Dr. Søren Nielsen M.D.

    Research Associate and Former Senior Registrar and Clinical Lecturer, Corresponding author
    1. Department of Child Psychiatry, Rigshospitalet, University of Copenhagen, Denmark
    • Kastelsvej 10st, DK-4000 Roskilde, Denmark
    Search for more papers by this author


A review of the literature on normal development and on psychiatric disorders related to seasonal variation, suggests that this phenomenon is probable to occur in anorexia nervosa. A search of the available literature is entirely negative with respect to seasonal variation in anorexia nervosa. Preliminary analysis suggests the existence of a season of onset, a season of birth, and also oi a season of admission in Danish female anorexia nervosa patients. A formally correct analysis for seasonal variation supports the existence of a season of onset in a selected sample, maximum is in May, amplitude 50%. A season of birth is found in an unselected nationwide sample of patients aged 15 years or older at first psychiatric admission for anorexia nervosa. Maximum late in April, amplitude 40% in the period 1963–1968. Season of admission could not be modelled as a sine wave, but the data suggested a pulse structure (a minimum in one month, followed by a maximum in the next month). Two annual pulses: December/January and July/August. Peak of onset is at a time of the year with maximum increase in height, and peak of other psychiatric conditions. Pattern of birth is similar to that reported for schizophrenia and mania. Replications of this preliminary study should be done in larger countries, at different latitudes and continents. Geophysical, demographic, and climatological variables should be controlled.