Two-hundred and fifty-three participants were invited to take part in the study. Forty-five declined (18%), 2 (1%) did not pass the cognitive impairment test, 23 (9%) did not meet the visual acuity eligibility criteria (<6/12 in the better eye), and visual acuity was unknown for a further 15 (6%). Six participants were later excluded because they were under 40-years old. Participants were more likely to have moderate or severe vision impairment compared with nonparticipants (P = 0.04). Participants did not differ significantly to nonparticipants in age or sex (both P > 0.05).
The final sample consisted of 162 adults with low vision (
Table 1). Participants mean age was 71.4 years (SD = 13.7, range, 42–94) and 67% of participants were female. The most prevalent eye diseases were AMD (40.1%), glaucoma (16%), and diabetic retinopathy (10.5%). Participants were more likely to have mild (<6/12–6/18) (38.3%) or moderate (<6/18–6/60) (40.1%) vision loss, rather than severe (<6/60) (21.6%). In terms of PHQ-9 thresholds for depressive symptoms, 77 (47.5%) of participants were experiencing minimal depressive symptoms (score of 0–4), 40 (24.7%) had mild depressive symptoms (score of 5–9), and 45 (27.8%) scored above the cut-off (score ≥ 10) point for moderate depressive symptoms. Thirty-one participants (19.1%) reported current use of antidepressants. The mean vision-specific distress score was 5.65 ± 1.85. Higher levels of depressive symptoms were strongly associated with lower vision-specific distress scores (i.e., greater vision-specific distress) (ρ = −0.694,
P < 0.001). A score of 4.93 on the vision-specific distress measure was predictive of the clinical cut-off point (score ≥ 10) on the PHQ-9. A score of 6.68 on the vision-specific distress measure indicated minimal depressive symptoms.
In univariate analysis (
Table 2) greater levels of vision-specific distress and depressive symptoms were both associated with: younger age, worse self-reported health, worse vision-specific functioning, lower perceived availability and adequacy of social support, greater use of avoidant coping, and poorer day-to-day, emotional and social coping efficacy (
P < 0.01). Both vision-specific distress and depressive symptoms were significantly related to history of depression (
P < 0.001). Those who reported to have experienced depression currently or in the past (
n = 69) had significantly higher PHQ-9 scores (mean = 9.45 ± 5.59) and greater levels of vision-specific distress (mean = 4.91 ± 1.89) compared with those who reported no previous experience of depression (
n = 92) (4.70 ± 4.97,
P < 0.001; 6.21 ± 1.62,
P < 0.001, respectively). Depressive symptoms and vision-specific distress were not associated with sex, marital status, or recruitment site (
P < 0.05).
Based on the four criteria for evaluating multivariable regression models, three best-fitting models for determining vision-specific distress were chosen (
Table 3). Each model accounted for just under 80% of the variance. Consistently across the models greater vision-specific distress was significantly associated with higher levels of depressive symptoms (e.g., model 1:
β = −0.06; confidence interval [CI]: −0.09, −0.02,
P = 0.001); poorer vision-specific functioning (
β = 0.64; CI: 0.51,0.77,
P < 0.001); greater avoidant coping (
β = −0.15; CI: −0.21, −0.10,
P < 0.001); and poorer efficacy in coping with the social aspects of vision impairment (
β = 0.32; CI:0.16,0.48,
P < 0.001). Commonality analysis indicated that vision-specific functioning accounted for 37.8% of unique variance in vision-specific distress (
Fig. 1). Depressive symptoms, avoidant coping, and efficacy of coping with the social challenges of vision impairment accounted for 22.8%, 22.4%, and 15.0% of the unique variance, respectively.
The three best-fitting multivariable models for determining depressive symptoms consistently included vision-specific distress as a variable significantly associated with greater depressive symptoms (e.g., model 1:
β = −1.51; CI: −1.89, −1.14,
P < 0.001;
Table 4). Model 1 also identified that those with a self-reported history of depression had higher levels of depressive symptoms (
β = 2.18; CI: 0.92, 3.44,
P = 0.001) and lower perceived adequacy of social support was associated with higher PHQ-9 scores (
β = −0.85; CI: −1.19, −0.51,
P < 0.001). Model 2 included all the variables shown in Model 1, in addition to health rating (
β = 0.66; CI: 0.05, 1.28,
P = 0.034) and avoidant coping (
β = .25; CI: 0.01, 0.49,
P = 0.043). Poorer self-reported health and greater avoidant coping were associated with higher levels of depressive symptoms. Model 3 included only vision-specific distress (
β = −1.11; CI: −1.62, −0.61,
P < 0.001) and avoidant coping (
β = 0.26; CI: 0.05, 0.47,
P = 0.015). Each model explained 56% to 58% of the variance in depressive symptoms. Commonality analysis showed that vision-specific distress accounted for 36.6% of the unique variance in depressive symptoms and avoidant coping explained 24.8% (
Fig. 2). Perceived adequacy of social support, self-reported history of depression and health rating accounted for 20.5%, 10.23%, and 7.9% of the unique variance, respectively.