Peer Effects on Mental Health and Health Behaviors: Evidence from College Roommates

Help-seeking and Utilization of Mental Health Care among Adolescents and Young AdultsDaniel Eisenberg, Ph.D.Department of Health Management and PolicyUniversity of Michigan (daneis@umich.edu)

Presentation for the UC Davis Center for Healthcare Policy and ResearchApril 15, 20131Disclosure StatementHave you (or your spouse/partner) had a personal financial relationship in the last 12 months with the manufacturer of the products or services that will be discussed in this CME activity?

___ Yes

_X_ No

2Educational objectives for this seminarDescribe data on mental health symptoms and utilization in college populations in the U.S.

Assess the economic case for interventions to increase help-seeking and access to mental health care in college and other youth populations

Discuss the effectiveness and potential effectiveness of specific interventions

3Outline of SeminarBroad overview of my work (5 minutes)

Help-seeking and utilization in college populationsGeneral statistics (10 minutes)Analysis of barriers to services (10 minutes)Economic case for access to services (10 minutes)Intervention research (10 minutes)

45

EconomicsPublic HealthPsychiatryClinical PsychologyEducation

PediatricsFamily MedicineBroad Research-Practice Agenda6How to invest most efficiently in health (and long-term success and wellbeing) in youth populations?Design and evaluate programs and interventionsCollect descriptive population dataPractice6Things I Like to DoEconomic evaluationCausal inference in nonexperimental settingsBridge between health and social sciences (not just economics)Bridge between health and education policyLarge-scale survey data collection (and data)Online interventions (access and self-efficacy)Training and mentoring junior scholars

7Opportunities for CollaborationEconomic analyses of policies, programs, servicesPopulation survey studiesBroad, preventive approaches to mental health and health behaviors through primary care settingsAddressing disparities (by race/ethnicity and SES) through school settingsOnline interventions: screening, linkage to health care, supplement to clinical care

89Help-seeking and Utilization of Mental Health Care in College Populations: General StatisticsSignificance of PopulationFor adolescents and young adults in the U.S., mental disorders account for the largest burden of disease 0f any type of health condition (Michaud et al, 2006, Pop Health Metrics)

75% of lifetime mental disorders in the U.S. have first onset by age 24 (Kessler et al, 2005, Arch Gen Psych)

Adolescence and young adulthood are periods of intensive investment in human capital

School settings offer unique opportunity for public health approaches with high impact10Why I think this is an interesting and important area, particularly from an economics perspective10Healthy Minds Study, 2007-2013

Finding #1: High Prevalence of Mental Health Problems, But also Positive Mental Health12Data: 2012 Healthy Minds Study (29 schools, ~25,000 survey respondents)Finding #2: About Half of Students with Mental Health Problems Receive Treatment13Data: 2012 Healthy Minds StudyAmong students with significant depressive symptoms and some treatment in past year, 57% received minimally adequate depression care (4+ psychotherapy visits or 2+ months of antidepressant medication)

Among all students with past-year depression, 22% received minimally adequate care

14Finding #3: When Provided, Depression Treatment is Less than Minimally Adequate in ~50% of CasesData: 2009 Healthy Minds StudyFinding #4a: Variation in Mental Health across Demographic Groups15Data: 2012 Healthy Minds StudyFinding #4b: Variation in Utilization across Demographic Groups16Data: 2012 Healthy Minds StudyFinding #5: Variation by Field of Study17Data: 2012 Healthy Minds StudyFinding #6: Risk/Protective FactorsRisk factors (negative correlation w/ mental health)Financial stress (both past and present)Experienced discrimination

Protective factors (positive correlation)Social supportReligiosityLiving on campus18Data: 2012 Healthy Minds Study19

Data: 2012 Healthy Minds StudyFinding #7: Variation across CampusesData: 2012 Healthy Minds Study20Help-seeking and Utilization of Mental Health Care in College Populations: Barriers to ServicesFindings on StigmaPersonal stigma low among college studentsOnly 12% of students agree with statement I think less of someone who has received MH treatment Perceived public stigma considerably higher64% agree with Most people think less of someone who has received MH treatmentPersonal stigma somewhat higher among: male, younger, Asian, international, religious, from a poor family21Stigma Findings (contd)Perceived public stigma not significantly associated with use of services or support

In contrast, personal stigma is significantly associated with lower use of services & supportOur estimates suggest that lowering the population-level personal stigma by one half would result in an increase of service use among students with major depression from 44% to 60%

22If Not Stigma, Then What?BARRIERS:stigma hightreatmentnot helpfulno perceived needN%Group 1XXX492%Group 2XX412%Group 3XX743%Group 4X472%Group 5XX34813%Group 6X32312%Group 7X86833%Group 889434%What is Going on with Groups 7 & 8?Group 7 (low stigma, believes tx helpful, no perceived need):prefer to deal with problems on ones own (53%)thinks stress is normal in school (47%)gets support from family/friends (42%)questions how serious issues are (36%)doesn't have time (29%)

Group 8 (low stigma, believes tx helpful, perceives need):questions how serious issues are (62%)prefers to deal with problems on ones own (60%)doesn't have time (59%)thinks stress is normal in school (59%)gets support from family/friends (44%)financial reasons (38%) 24Interventions for Groups 7 & 8?Anti-stigma, education, and awareness campaigns may have little impact

May be useful to borrow lessons from other contexts where people do not have strong objections, yet fail to engage in healthy behaviors (e.g., exercise, diet, preventive screening, even saving for retirement!)25Behavioral Economics: Time Preferences and ProcrastinationIs depression related to present-orientation (discounting of future)?

Is lack of help-seeking a form of procrastination?26Empirical Analysis of these QuestionsHealthy Minds Study (2011)Large, cross-sectional (N=8,806, 11 institutions)College Transition Study Replication (CTSR)Panel with five monthly surveys (Aug-Dec 2010) at one institution (Univ. Michigan)281 first-year and transfer undergraduatesPI: Steve Brunwasser

27FindingsDepressive symptoms significantly associated with present-orientation (discounting the future) and procrastination tendencies

Procrastination tendencies associated with lower likelihood of receiving treatment

Implications for help-seeking interventions?2829Help-seeking and Utilization of Mental Health Care in College Populations: Economic CaseMental Health and Academic OutcomesMental health as predictor of academic outcomes in 2005-2008 Healthy Minds data

Depression (PHQ-9 score) is a significant predictor of dropping out10 point lower PHQ-9 scorereduction in risk of dropping out by a multiple of 0.6 (e.g., from 10% to 6%)Presenter: Daniel Mental Health and Grade Point Average (GPA)Depression (PHQ-9 score) is also a significant negative predictor of same-semester GPA10 point lower PHQ-9 score 9 point increase in GPA percentile

Co-occurrence of depression and anxiety associated with a significant additional drop in GPA.

Symptoms of eating disorders also associated with lower GPAPresenter: Daniel Economic Case for Services and Programs for Student Mental HealthReduced depressionIncreased retentionIncreased tuitionIncreased lifetime productivity (earnings)Increased student satisfactionIncreased institutional reputation & alumni donationsBenefits to institutionBenefits to students and societyPresenter: Daniel 3233Help-seeking and Utilization of Mental Health Care in College Populations: Intervention ResearchGatekeeper Training ProgramsEvaluation of Mental Health First Aid training for resident advisors (RAs)Co-PIs: Daniel Eisenberg and Nicole SpeerFunder: NIMH (2009-2011)

32-campus randomized trial to assess impacts on student communities

34Peer-based Approaches to Help-seekingPeer effects in mental health among college studentsPI: Daniel Eisenberg (University of Michigan)Funder: W.T. Grant Foundation (2009-2011)

Study design based on natural experiment of randomly assignment of students to roommates and resident advisors (RAs)35Online Screening and Linkage to Treatmente-Bridge to Mental Health online interventionPI: Cheryl King (University of Michigan)Funder: NIMH (2009-2012)

Brief risk screen -> personalized feedback -> correspondence with counselor using motivational interviewing36Online Video-based InterventionBrief (3-4), highly engaging videos based on CBT and resilience and self-efficacy skills

Based on inkblots video series (www.inkblots.tv)

Pilot RCTs to begin in summer 2013 (funded by UM Comprehensive Depression Center)37Broad Research-Practice Agenda38How to invest most efficiently in health (and long-term success and wellbeing) in youth populations?Design and evaluate programs and interventionsCollect descriptive population dataPractice38