outcome,unmeasured_confounder_RR_to_explain_away,unmeasured_confounder_RR_to_shift_CI_to_null,num_longitudinal_studies,pct_showing_protective,evidence_grade,source
All-cause mortality,2.35,1.67,69+ (meta-analysis),majority,Strong,Li et al 2016; VanderWeele sensitivity analysis
Suicide,very strong (HR=0.16 base),very strong,handful - low base rate,all,Strong,VanderWeele et al 2016; sensitivity analysis
Depression,1.74,1.46,20+ with baseline control,77%,Strong,"Garssen et al 2021; Balboni et al 2022 systematic review; VanderWeele 2021"
Substance use disorders,N/A,N/A,numerous,majority,Moderate-Strong,"Weinberger et al 2022; multiple reviews"
Life satisfaction,N/A,N/A,mostly cross-sectional,majority,Moderate,Lim & Putnam 2010; Chen et al 2020 longitudinal
Loneliness,N/A,N/A,several longitudinal,majority,Moderate,Rote et al 2013; Chen et al 2020
Anxiety,N/A,N/A,limited,mixed,Weak-Mixed,Chen et al 2020; various
Purpose in life,N/A,N/A,few longitudinal,majority,Moderate,Chen et al 2020
Physical disease (hypertension etc),N/A,N/A,few,null results,Null,Chen et al 2020 - attendance does NOT predict incident disease
Sexual minority populations,N/A,N/A,limited,reversed for attendance,Context-dependent,"NSDUH 2016-2019; attendance INCREASES suicide risk for LGBT"