Estimating peer effects in longitudinal dyadic data using instrumental variables.

The identification of causal peer effects (also known as social contagion or induction) from observational data in social networks is challenged by two distinct sources of bias: latent homophily and unobserved confounding. In this paper, we investigate how causal peer effects of traits and behaviors can be identified using genes (or other structurally isomorphic variables) as instrumental variables (IV) in a large set of data generating models with homophily and confounding. We use directed acyclic graphs to represent these models and employ multiple IV strategies and report three main identification results. First, using a single fixed gene (or allele) as an IV will generally fail to identify peer effects if the gene affects past values of the treatment. Second, multiple fixed genes/alleles, or, more promisingly, time-varying gene expression, can identify peer effects if we instrument exclusion violations as well as the focal treatment. Third, we show that IV identification of peer effects remains possible even under multiple complications often regarded as lethal for IV identification of intra-individual effects, such as pleiotropy on observables and unobservables, homophily on past phenotype, past and ongoing homophily on genotype, inter-phenotype peer effects, population stratification, gene expression that is endogenous to past phenotype and past gene expression, and others. We apply our identification results to estimating peer effects of body mass index (BMI) among friends and spouses in the Framingham Heart Study. Results suggest a positive causal peer effect of BMI between friends.

The spread of alcohol consumption behavior in a large social network.

BACKGROUND: Alcohol consumption has important health-related consequences and numerous biological and social determinants. OBJECTIVE: To explore quantitatively whether alcohol consumption behavior spreads from person to person in a large social network of friends, coworkers, siblings, spouses, and neighbors, followed for 32 years. DESIGN: Longitudinal network cohort study. SETTING: The Framingham Heart Study. PARTICIPANTS: 12 067 persons assessed at several time points between 1971 and 2003. MEASUREMENTS: Self-reported alcohol consumption (number of drinks per week on average over the past year and number of days drinking within the past week) and social network ties, measured at each time point. RESULTS: Clusters of drinkers and abstainers were present in the network at all time points, and the clusters extended to 3 degrees of separation. These clusters were not only due to selective formation of social ties among drinkers but also seem to reflect interpersonal influence. Changes in the alcohol consumption behavior of a person's social network had a statistically significant effect on that person's subsequent alcohol consumption behavior. The behaviors of immediate neighbors and coworkers were not significantly associated with a person's drinking behavior, but the behavior of relatives and friends was. LIMITATIONS: A nonclinical measure of alcohol consumption was used. Also, it is unclear whether the effects on long-term health are positive or negative, because alcohol has been shown to be both harmful and protective. Finally, not all network ties were observed. CONCLUSION: Network phenomena seem to influence alcohol consumption behavior. This has implications for clinical and public health interventions and further supports group-level interventions to reduce problematic drinking.

The impact of poor health on academic performance: New evidence using genetic markers.

This paper examines the influence of health conditions on academic performance during adolescence. To account for the endogeneity of health outcomes and their interactions with risky behaviors we exploit natural variation within a set of genetic markers across individuals. We present evidence that specific genetic markers have good statistical properties to identify the impacts of ADHD, depression and obesity. These markers help reveal a new dynamism from poor health to lower academic achievement with substantial heterogeneity in their impacts across genders. Our investigation further exposes the considerable challenges in identifying health impacts due to the prevalence of comorbid health conditions, with clear implications for the health economics literature.

Validation of electronic health record phenotyping of bipolar disorder cases and controls.

OBJECTIVE: The study was designed to validate use of electronic health records (EHRs) for diagnosing bipolar disorder and classifying control subjects. METHOD: EHR data were obtained from a health care system of more than 4.6 million patients spanning more than 20 years. Experienced clinicians reviewed charts to identify text features and coded data consistent or inconsistent with a diagnosis of bipolar disorder. Natural language processing was used to train a diagnostic algorithm with 95% specificity for classifying bipolar disorder. Filtered coded data were used to derive three additional classification rules for case subjects and one for control subjects. The positive predictive value (PPV) of EHR-based bipolar disorder and subphenotype diagnoses was calculated against diagnoses from direct semistructured interviews of 190 patients by trained clinicians blind to EHR diagnosis. RESULTS: The PPV of bipolar disorder defined by natural language processing was 0.85. Coded classification based on strict filtering achieved a value of 0.79, but classifications based on less stringent criteria performed less well. No EHR-classified control subject received a diagnosis of bipolar disorder on the basis of direct interview (PPV=1.0). For most subphenotypes, values exceeded 0.80. The EHR-based classifications were used to accrue 4,500 bipolar disorder cases and 5,000 controls for genetic analyses. CONCLUSIONS: Semiautomated mining of EHRs can be used to ascertain bipolar disorder patients and control subjects with high specificity and predictive value compared with diagnostic interviews. EHRs provide a powerful resource for high-throughput phenotyping for genetic and clinical research.

Seasonality in seeking mental health information on Google.

BACKGROUND: Population mental health surveillance is an important challenge limited by resource constraints, long time lags in data collection, and stigma. One promising approach to bridge similar gaps elsewhere has been the use of passively generated digital data. PURPOSE: This article assesses the viability of aggregate Internet search queries for real-time monitoring of several mental health problems, specifically in regard to seasonal patterns of seeking out mental health information. METHODS: All Google mental health queries were monitored in the U.S. and Australia from 2006 to 2010. Additionally, queries were subdivided among those including the terms 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 isolate seasonal components in the trends, and based on this model, the mean search volume in winter was compared with that in summer, as performed in 2012. RESULTS: All mental health queries followed seasonal patterns with winter peaks and summer troughs amounting to a 14% (95% CI=11%, 16%) difference in volume for the U.S. and 11% (95% CI=7%, 15%) for Australia. These patterns also were evident for all specific subcategories of illness or problem. For instance, seasonal differences ranged from 7% (95% CI=5%, 10%) for anxiety (followed by OCD, bipolar, depression, suicide, ADHD, schizophrenia) to 37% (95% CI=31%, 44%) for eating disorder queries in the U.S. Several nonclinical motivators for query seasonality (such as media trends or academic interest) were explored and rejected. CONCLUSIONS: Information seeking on Google across all major mental illnesses and/or problems followed seasonal patterns similar to those found for seasonal affective disorder. These are the first data published on patterns of seasonality in information seeking encompassing all the major mental illnesses, notable also because they likely would have gone undetected using traditional surveillance.