Table 2.

Characteristics and Findings of Included Studies of Hearing Loss and Social Isolation.

Study	Sample	Hearing Assessment	Loneliness Assessment	Results	Study Strengths, Limitations, and Quality
Kramer et al (2002)25	N = 3107 participants recruited from the Longitudinal Aging Study Amsterdam, a community-based random sample; age range 55–85 y	Self-report: scaled score based on responses to 3 questions about hearing conversation and telephone with and without hearing aids, dichotomized for analysis	Size of the social network (7 domains of network members with whom respondent maintained a regular relationship)	Hearing loss associated with smaller network (12.7 vs 14.1, P <.01); in multiple linear regression, hearing impairment associated with lower network size (regression coefficient = −0.52, P <.01)	Strengths: large sample size, random sampling of underlying source population NOS score: 5 (low risk of bias)
Mick et al (2014)33	N = 1453 participants from the 1999 to 2006 cycle of the NHANES, a community-based random sample; age range 60–84 y	Pure-tone audiometry to measure PTA at 0.5, 1, 2, and 4 kHz	A summary social isolation score (SIS) based on the SSQ	Among 60- to 69-y-olds, odds of social isolation was 2.14 (95% CI, 1.29–3.57) higher per 25 dB of hearing loss; there was an interaction with gender (women: OR 3.49; 95% CI, 1.91–6.39; men: OR: 1.11; 95% CI, 0.66–1.88); among 70- to 84-y-olds, there was no significant increase in odds of social isolation with each 25-dB increase in hearing loss (OR 1.24; 95% CI, 0.75–2.0)	Strengths: random sample of participants, audiometric measure of hearing Limitations: NHANES SSQ questionnaire and SIS have not been standardized NOS score: 5 (low risk of bias)
Mick et al (2018)36	N = 21,241 community-dwelling older adults age 45–89 y, a nationally representative sample from the Canadian Longitudinal Study of Aging	Self-reported hearing status based on the following question: “Is your hearing, using a hearing aid if you use one: Excellent, Very Good, Good, Fair, Poor or Nonexistent or deaf”	Social network diversity determined using a modified version of the SNI; availability of social support measured using the Medical Outcomes Social Support Survey	Hearing loss was not associated with social network diversity (mean difference in SNI score between those with hearing loss and those without = 0.02, 95% CI, 0.06, 0.10); hearing loss was associated with having a social support score lower than the median (OR 1.22; 95% CI, 1.10, 1.25)	Strengths: large sample size, nationally representative sample; used multiple validated measures of social isolation and social support NOS score: 5
Mick and Pichora-Fuller (2016)34	N = 1820 participants from the 1999 to 2010 cycles of NHANES	Self-reported hearing status, hearing test history, and hearing aid use; pure-tone audiometry; participants categorized as having unacknowledged hearing loss (PTA >25 dB but self-reported normal hearing) and as having unaddressed hearing loss (PTA >25, self-reported hearing difficulty but no hearing test or hearing aid use)	A summary SIS based on SSQ: not married or in a domestic partnership, no close friends, no one to provide financial support, no one to provide emotional support; participants were considered isolated if they met 2 of the above criteria	A 10-dB increase in PTA was associated with 1.52 higher odds of social isolation among 60- to 69-yolds with unacknowledged or unaddressed hearing loss (OR: 1.52; 95% CI, 1.19–1.93); there was no association between higher odds of social isolation with each 10-dB increase in PTA among 701 y olds (OR 1.08; 95% CI, 0.77–1.52)	Strengths: random sample of participants, audiometric measure of hearing Limitations: NHANES SSQ questionnaire and SIS have not been standardized; did not look at participants with treated hearing difficulty NOS score: 5
Weinstein and Ventry (1982)32	N = 80 male veterans recruited through outpatient centers at a Veterans Affairs medical center; age range 65–88 y, mean age 74 y	Pure-tone audiometry; hearing handicap using the HMS; speech detection test using the W-22 PB work list and the Rush-Hughes PB-50 word list	Objective Isolation Scale from the Comprehensive Assessment and Referral Evaluation (CARE) Questionnaire	Objective isolation was positively correlated with all measures of hearing: HMS (r = 0.26), PTA (r = 0.24), Rush-Hughes word list (r = 0.22), and W-22 word list (r = 0.18	Limitations: male veterans only, findings nongeneralizable; unadjusted correlations between hearing and loneliness with no control for relevant confounders NOS score: 1

Abbreviations: CI, confidence interval; HMS, Hearing Measurement Scale; NHANES, National Health and Nutrition Examination Survey; OR, odds ratio; PTA, pure-tone average; SIS, social isolation score; SNI, Social Network Index; SSQ, Social Support Questionnaire.