The objective of this study is to determine whether human body odors undergo seasonal modulation. We utilized google trends search volume from the United States of America from January 1, 2010 to June 24, 2017 for a number of predetermined body odors. Regression modeling of time series data was completed. Our primary outcome was to determine the proportion of the variability in Internet searches for each unpleasant odor (about the mean) that is explained by a seasonal model. We determined that the seasonal (sinusoidal) model provided a significantly better fit than the null model (best straight line fit) for all searches relating to human body odors (P <.0001 for each). This effect was easily visible to the naked eye in the raw time series data. Seasonality explained 88% of the variability in search volume for flatulence (i.e. R2 = 0.88), 65% of the variability in search volume for axillary odor, 60% of the variability in search volume for foot odor, and 58% of the variability in search volume for bad breath. Flatulence and bad breath tended to peak in January, foot odor in February, and Axillary odor in July. We conclude that searching by the general public for information on unpleasant body odors undergoes substantial seasonal variation, with the timing of peaks and troughs varying with the body part involved. The symptom burden of such smells may have a similar seasonal variation, as might the composition of the commensal bacterial microflora that play a role in creating them.
Keywords: microbiome; odor; seasonality; smell.