Novel Identification of Seasonality in Mental Health Problem Inquiries

Because mental health surveillance lacks the temporal resolution to make comparisons across seasons, seasonality is under-explored. By monitoring aggregate Internet search queries for several mental health problems, we detected seasonal variability. Millions of health-related Internet search queries may be aggregated to form a continuous and real-time data stream. All mental health queries were monitored for 2006-2010 in the United States, cross-validated in Australia, with additional monitoring among 72 cities. Queries were also sub-divided among those containing: “ADHD” (attention deficit-hyperactivity disorder), “anxiety,” “bipolar”, “depression”, “anorexia” or “bulimia” (eating disorders), “OCD” (obsessive compulsive disorder), “schizophrenia,” and “suicide.” A wavelet phase analysis was used to identify seasonal patterns. All mental health queries followed seasonal patterns, declining 15% (95%CI,13-17) during the summer (June, July, and August) compared to the remainder of the year in the US, and 14% (95%CI,11-18) in Australia (December, January, and February). These patterns were consistent for all problem-specific outcomes, ranging from 5% (95%CI,3-6) for anxiety to 25% (95%CI,23-28) for ADHD in the US, for example. City-level analysis suggested amplitudes of summer decline were stronger moving further from the equator for depression, schizophrenia and eating disorder queries but not ADHD, bipolar or anxiety queries. A major issue in mental health epidemiology is how to make population assessments; Internet search queries may be one such approach. Hypothesized explanations for summer declines in mental health inquiries likely include biologic and social pathways. These trends may make clinical resource delivery more cost-effective by reaching populations when they are considering their mental health.