The Impact of Diabetic Retinopathy on Participation in Daily Living

Objective  To determine the restriction of participation in daily activities ofpeople with diabetic retinopathy using the Impact of Vision Impairment questionnaire.

Methods  Individuals with diabetic retinopathy and a visual acuity (VA) worsethan 20/40 or 6/12 in the better eye were eligible. Participants answereddemographic questions and had VA information abstracted from medical records.If VA information was unavailable, it was assessed by an orthoptist.

Main Outcome Measures  All participants completed the Impact of Vision Impairment questionnaire,which was either self-administered or interviewer administered. The physicaland mental health components were assessed using the Medical Outcomes Study12-Item Short Form (SF-12) questionnaire.

Results  Forty-five participants (mean age, 67.5 years) were recruited, withalmost 70% (30/45) recording a VA worse than 20/60 or 6/18 in the better eye.The median duration of vision loss was 2.0 years. The highest restrictionwas reported for the Leisure and Work, Mobility, and Consumer and Social Interactiondomains (mean, 3.0, 2.8, and 2.8, respectively), compared with the EmotionalReaction to Visual Loss and Household and Personal Care domains (mean, 2.3and 2.1, respectively) (P<.005). The activitieswith the greatest restriction of participation were reading print, mobility,work, and leisure. A poorer VA in the better eye correlated independentlywith increased restriction of participation, as measured by the Impact ofVision Impairment questionnaire scores (partial correlations, 0.29-0.41; P≤.03).

Conclusion  Low-vision rehabilitation services aiming to improve outdoor mobility,print reading, participation in leisure activities, and psychological healthmay be an effective strategy to help people with diabetic retinopathy increasetheir participation in daily activities.

Diabetic retinopathy (DR) is an important cause of visual loss in adults,with significant medical, social, and financial consequences.¹ Theprevalence of DR among people with diabetes mellitus in Australia and theUnited States is between 22% and 36%.^2-6 Abest-corrected visual acuity (VA) of 20/200 or worse has been estimated tobe 25 times more prevalent in the diabetic population than in those withoutthe disease.^7-9 Visionimpairment resulting from DR can, however, be prevented with regular screeningand timely treatment.¹⁰

Vision impairment has been linked with dependency in activities of dailyliving,^11-13 socialisolation,¹⁴ and reduced physical activity.¹⁵ Several studies^16-19 havealso examined the relationship between different measures of vision and disability.However, the magnitude of restriction of participation in daily activitiesas a consequence of vision impairment has not been extensively investigated.Restriction of participation, previously known as handicap, is the limitationdue to an impairment or disability on activities that an individual needsor wants to perform.²⁰ By assessing restrictionof participation, it is possible to design and implement strategies to effectivelyincrease participation and consequently improve the quality of life of visuallyimpaired people. Recently, the Centre for Eye Research Australia developedand validated a questionnaire that can measure a person's restriction of participationin 32 items. The Impact of Vision Impairment (IVI) questionnaire has beenshown to be valid and reliable,²¹ but has notyet been used to measure the impact of the restriction of participation ofa specific eye disease in the general population.

This investigation, therefore, quantifies and describes the restrictionof participation in daily activities subsequent to vision impairment in peoplewith DR using the IVI questionnaire.

Participants were recruited from the Vision Australia Foundation (VAF;Victoria, Australia), a low-vision service provider, and The Royal VictorianEye and Ear Hospital (RVEEH), East Melbourne. Only participants with no historyof vision rehabilitation were included. Other criteria for inclusion werea VA worse than 20/40 or 6/12 in the better eye, a diagnosis of DR as themain condition causing vision loss, the ability to converse in English, andbeing 18 years or older. All participants were informed of the nature andpurpose of the study, and gave witnessed consent. All participants answereddemographic questions regarding age, duration of vision impairment, hearingloss, other health conditions, and limitations on performing daily activitiesattributed to other health conditions. Participants completed the IVI questionnaire.The study received ethics approval from the RVEEH and VAF in accordance withthe Declaration of Helsinki for research involving human subjects.

Participants underwent eye examinations that included measurement ofVA, and underwent refraction as a routine part of their appointments at theRVEEH and the VAF. If VA information was unavailable, qualified personnelassessed it. The best-corrected distance VA and near VA in the better eyewere abstracted from the files for the participants. The distance VA was categorizedas mild (<20/40 to 20/60 [<6/12 to 6/18]), moderate (<20/60 to 20/200[<6/18 to 6/60]), and severe (<20/200 [<6/60]). The near VA was categorizedas N8 or better, less than N8 to N20, less than N20 to N48, and less thanN48.

A detailed description of the IVI questionnaire has been fully publishedelsewhere.²¹ Briefly, the IVI questionnairecomprised 32 items grouped under 5 domains (Leisure and Work, Consumer andSocial Interaction, Household and Personal Care, Mobility, and Emotional Reactionto Vision Loss). Participants with a VA sufficient to read large print (18-pointfont) self-administered the questionnaire before their first visit at thelow-vision clinics. Otherwise, the 32-item IVI questionnaire was intervieweradministered at the participant's first visit at the RVEEH and the VAF. Theinterviewer (J.B.H.) was trained to administer the IVI questionnaire. Proxyanswers were not solicited from caregivers or relatives to avoid biasing theIVI questionnaire responses to the perception of another person's opinionof the participant's ability. High levels of consistency have been shown betweenself- and interview-administered methods for the IVI questionnaires.²¹

Responses to the IVI questionnaire items under the domains were ratedas follows: "not at all" (0), "rarely" (1), "a little" (2), "a fair amount"(3), "a lot" (4), "can't do because of eyesight" (5), or "don't do becauseof other reasons" (8). Responses to the items under the Emotional Reactionto Vision Loss domain were rated as follows: "not at all" (0), "very rarely"(1), "a little of the time" (2), "a fair amount of the time" (3), "a lot ofthe time" (4), "all of the time" (5), or "don't do because of other reasons"(8). Data in parentheses are scores. Total and domain scores of the IVI questionnaireare an arithmetic average of items rated between 0 (the best score) and 5(the worst score). An item rated with a score of 8 was not included in thefinal average score, and was analyzed separately.

Participants also completed the Medical Outcomes Study 12-Item ShortForm (SF-12; version 1: Physical and Mental Health Summary Scales). The SF-12(a short validated version of the Medical Outcomes Study 36-Item Short-FormHealth Survey) was included to evaluate the physical and mental health componentsof the participants. It was also used to determine if the overall health ofthe participants was a potential confounder when assessing the relationshipbetween IVI questionnaire score and other variables of interest. By usingthe algorithm developed by Ware and associates,²² 2summary components related to the physical (Physical Summary Scale) and mental(Mental Summary Scale) domains of life were computed from the questions inthe SF-12. Each summary scale is scored from 0 to 100, where 100 indicatesthe best possible score and 0 represents the worst possible score.

Descriptive analyses were performed to characterize the sociodemographic,health, and clinical characteristics of the study participants. Spearman rankcorrelation tests were performed to determine the association between theIVI questionnaire scores and the participants' demographic and clinical characteristics.The Mantel-Haenszel test was used to determine the association between VAand the IVI questionnaire domain and overall scores. The partial correlationsprocedure was used to compute partial correlation coefficients between VAand the IVI questionnaire scores while controlling for age, sex, the durationof eye impairment, comorbidity, Physical Summary Scale score, and Mental SummaryScale score. The Wilcoxon rank sum tests were selected to compare differencesin the distributions of IVI questionnaire domains, and the χ² testswere used to compare proportions. An α level of P<.05 was chosen as the criterion for significance for all the statisticaltests, except for the Wilcoxon rank sum test, which was set at P<.005 because of the many paired comparisons undertaken.

Forty-five subjects (20 men) with DR as their main cause of vision lossparticipated in the study (Table 1).Twenty-seven subjects were clients of the VAF, and the remaining 18 were patientsof the RVEEH. More than 53% (24/45) of the participants were born in Australia,and nearly 80% (35/45) used English as the main spoken language at home. Allparticipants had diabetes mellitus and other diabetic-related health complications,such as hypertension, a heart condition, gout, and asthma. Of the participants,42% (19/45) reported that other health conditions interfered "a great deal"with their activities. Only 6 subjects were employed either part time or fulltime, and they were aged between 47 and 59 years.

Almost 70% of the participants had a distance VA worse than 20/60 inthe better eye when they were first seen (Table 1). Following refraction, 56% had a distance VA worse than20/60. Only 40% of the participants recorded a score of N8 or better. Themedian duration of vision impairment was 2.0 years (range, 0-33 years). Almost18% (8/45) of the participants reported having their vision affected for 10years or longer.

The mean domain and overall scores were greater than 2, indicating thaton average the participants ranked the items in the domains between a littleand a fair amount of restriction. The highest scores were recorded for theLeisure and Work, Mobility, and Consumer and Social Interaction domains (meanscores, 3.0, 2.8, and 2.8, respectively), compared with the Emotional Reactionto Vision Loss and Household and Personal Care domains (mean scores, 2.3 and2.1, respectively) (Figure 1) (Wilcoxonrank sum test, P<.005).

The mean ± SD score of the individual IVI questionnaire itemswas 2.6 ± 1.1 (range, 1.4-4.0) (Table 2). When ranked from the most restrictive items, 11 itemsshowed that the participants experienced between a fair amount and a lot ofrestriction in these activities (mean scores, 3-4) (Table 2). Of the items with the greatest magnitude of restriction,3 were from the Mobility and Emotional Reaction to Vision Loss domains eachand 2 were from the Leisure and Work and Consumer and Social Interaction domainseach. Most of the items of the Household and Personal Care domain caused littlerestriction for the study participants (Table 2).

The reading of small print, labels, and street signs caused the greatestrestriction because these activities were among the items with the highestmean scores (Table 2). Participationin paid work or leisure activities was also rated as fairly restrictive, withmean scores ranging between 2.8 and 3.4. Three items with a high degree ofrestriction were associated with mobility (public transport use, obstructedlocomotion [eg, stairs and curbs], and fear of falling) (mean score, 3.0).The participants also highly rated 3 emotional health–related itemsand stated that for a fair amount to a lot of the time they were concernedbecause their eyesight was getting worse and they felt frustrated and likea nuisance because of their DR. On the other hand, the participants did notfeel overly sad, depressed, or lonely because of their eyesight, because these3 items from the Emotional Reaction to Vision Loss domain scored between 1.4and 1.8.

There was at least one category response of cannot do because of eyesightfor each item of the IVI questionnaire (Table 2). Eight items recorded at least 20% of the total responsesas cannot do because of eyesight. These items were "Reading ordinary sizeprint," "Reading a sign across the street," "Reading labels or instructionson medicines," "Favorite pastimes or hobbies," "Worried because eyesight gettingworse," "Paid or voluntary work," "Eyesight interfered with using transport,"and "Frustrated or annoyed because of your eyesight." No significant differenceswere found in the proportions of those who reported cannot do because of eyesightand the other responses for age, sex, near and distance VA, and duration ofvision loss (χ² test, P>.05 for all).

No significant statistical correlations (Spearman rank correlation, P>.05) were found between VA and age, sex, level of education,and duration of visual impairment. On the other hand, a poorer distance VAin the better eye significantly correlated with the increased restrictionin the IVI questionnaire domain scores (Mantel-Haenszel test, 5.9-7.9; P≤.02) (Figure 2).The association was still evident after controlling for age, sex, durationof vision loss, comorbidity, Physical Summary Scale score, and Mental SummaryScale score (partial correlations, 0.29-0.41; P range,.01-.03).

Our participants reported that the functional life domains causing thegreatest restriction of participation because of their DR were Leisure andWork, Mobility, Consumer and Social Interaction, and the Emotional Reactionto Vision Loss, with 19 of the 20 most restrictive items originating fromthese 4 domains. The Household and Personal Care domain, on the other hand,was the least restrictive. Between a moderate and a lot of restriction ofparticipation was reported for activities associated with print reading, work,leisure, the use of public transport, and outdoor mobility. Our results showthat people with DR experience at least a moderate amount of restriction ofparticipation in a range of daily activities and life domains fundamentalto a good quality of life.

Our domain and individual rankings were similar to those of 2 studiesthat also used the IVI questionnaire, namely, those of Hassell et al,²³ using patients with glaucoma, and Weih et al,²¹ who recruited participants with age-related maculopathy,glaucoma, DR, cataract, and other retinopathies. Collectively, these findingsdemonstrate the relevance of the IVI questionnaire as a meaningful questionnaireto identify the nature and magnitude of restriction of participation in dailyactivities for visually impaired individuals.

The Household and Personal Care domain low ranking is not surprisingconsidering that familiarity with the household environment can make itemslike operating household appliances and not spilling and breaking things easierto undertake compared with items under the other domains. The distributionof the items under the Emotional Reaction to Vision Loss domain was also interesting.Our participants did not feel depressed or isolated, because items such as"Felt sad or low because of eyesight," "Embarrassed because of eyesight,"and "Felt lonely or isolated because of eyesight" were lowly ranked. However,items such as "Worried because eyesight getting worse" and "Felt like a nuisanceor burden because of eyesight" were ranked highly, with a mean score of 3.1.The high ranking of these items suggests that while the emotional domain recordedan overall mean score of 2.3, some items of this domain have a substantialamount of restriction of participation but may be masked by some relativelylow scores of other items of the same domain. This finding suggests that despiteour small sample size and the modest overall score of the emotional domainrecorded in this study, a psychological component seems to be an essentialaspect of vision rehabilitation strategies for individuals with DR.

Significant modest correlations were found between the IVI questionnairescore and VA in the better eye. In addition, a closer scrutiny of Figure 2 seems to show that participantswith severe vision impairment (VA, <20/200) experience greater restrictionof participation on the IVI questionnaire items than those with mild and moderatevision impairments (VA, <20/40 to 20/60 and 20/60 to 20/200, respectively).This observation suggests that eye care providers, when treating a severelyvision-impaired individual, should inquire about participation in the essentialdomains of life. This finding also confirms that despite the relatively smallsample size of the present study, the IVI questionnaire is a relevant andresponsive instrument for assessing participation in vision-impaired individualswith different levels of VA.

The mean age of our participants was 68 years, and a third were youngerthan 60 years. While most eye conditions are age related, our sample withDR contained several relatively younger participants, probably because ofthe young age of onset of diabetes mellitus. Young persons whose conditionis due to DR have also been observed among participants aged 50 to 59 yearsin the Baltimore Eye Study²⁴ and the MelbourneVisual Impairment Project.²⁵ More important,the mean overall IVI questionnaire score of our younger subjects was 3.1 (vs2.3 for participants older than 60 years; P = .07),suggesting that the younger participants are experiencing similar restrictionof participation. These data suggest that age should not be used as a guideto provide vision rehabilitation services to individuals with DR. Supportfor this suggestion is further provided because no significant correlationwas found between age and the IVI questionnaire scores.

Finally, the findings of the present investigation should be viewedwithin 2 limitations. First, our sample size was relatively small becauseof the specific criteria restricting our participants' selection. Based onour study sample of 45 participants, the average level of participation scoreof 2.6 was estimated (95% confidence interval, 2.2-2.9). Second, we did notinclude a comparison group. Consequently, future investigations with largersample sizes and comparison groups are required to confirm our findings. Inaddition, our sample was limited to participants who could converse in English,and the findings of this study should not be extended to the sections of theAustralian non–English-speaking population.

In conclusion, the IVI questionnaire was designed to determine the degreeof restriction of participation for people with impaired vision to provideeffective rehabilitation strategies to increase participation. Based on ourfindings, low-vision rehabilitation services with programs aiming to improveoutdoor mobility, print reading, participation in leisure activities, andpsychological health could be an effective strategy to help individuals withDR increase their participation in activities of daily living.

Corresponding author: Ecosse L. Lamoureux, PhD, Centre for Eye ResearchAustralia, Department of Ophthalmology, University of Melbourne, Locked Bag8, East Melbourne, Victoria 8002, Australia (e-mail: ecosse@unimelb.edu.au).

Submitted for publication December 4, 2002; final revision receivedAugust 24, 2003; accepted September 10, 2003.