Development of a Questionnaire to Assess Vision Problems under Low Luminance in Age-Related Maculopathy

Abstract

purpose. To develop a questionnaire for assessing self-reported visual problems under low luminance and at night for use in studies on age-related maculopathy (ARM).

methods. The questionnaire was developed in three steps: (1) Content for questionnaire items was identified through focus groups of older adults with ARM and those exhibiting normal retinal aging. The topic for discussion was “vision at night and under low lighting.” Discussion was audiotaped, transcribed, and subjected to content analysis to identify problem categories expressed by patients. (2) This content was used to develop a preliminary questionnaire administered by telephone to persons with ARM or normal retinal aging. Principal-components analysis identified groups of items that formed the questionnaire’s subscales that were evaluated for internal consistency, and an item-reduction strategy was implemented to generate a briefer questionnaire. (3) Psychometric properties of the shortened Low-Luminance Questionnaire (LLQ) were determined, including construct validity, criterion validity, and test–retest reliability.

results. The 32-item LLQ has six subscales (driving, extreme lighting, mobility, emotional distress, general dim lighting, and peripheral vision), all with good internal consistency (Cronbach α ≥ 0.82). Scores on LLQ subscales correlated moderately with nearly all National Eye Institute Visual Function Questionnaire (NEI VFQ)-25 subscales and decreased in value (indicating more disability) for patients with increasing ARM disease severity. Whereas rod-mediated parameters of dark adaptation were significantly associated with LLQ subscale scores (r = 0.19–0.43, all P < 0.03), cone-mediated parameters were not. Test–retest reliability ranged from 0.74 to 0.88 for all subscales (P < 0.0001), except for peripheral vision (0.46; P = 0.0003), which also exhibited a ceiling effect in almost half of the respondents.

conclusions. The 32-item LLQ, derived from the content of focus group comments by persons with ARM, has good construct validity, subscale scores related to rod-mediated visual function, and good test–retest reliability for five of six subscales. The LLQ may ultimately be useful in patient-centered evaluation of the outcome of interventions to prevent ARM or to arrest progression of early disease.

Among the common visual symptoms of age-related maculopathy (ARM) are problems seeing in dim lighting and at night. ^{1 2 3} Even in the early phases of disease when visual acuity is unimpaired, these symptoms are present and associated with rod-mediated sensitivity impairment and dark adaptation delays, thus suggesting a functional basis. ^{1 4} These findings are consistent with histopathological findings in donor tissue that show that rod photoreceptors are selectively vulnerable in ARM. Rod photoreceptors show the earliest signs of degeneration in most ARM donor eyes studied, their degeneration is more severe than for cones at all stages of disease, and the last surviving photoreceptors appear to be cones. ^{5 6 7} There also have been psychophysical demonstrations of rod vulnerability over cones, particularly early in the disease. ^{4 8 9 10}  

Existing questionnaires may be inadequate for assessing task difficulty or problems encountered under low luminance. The commonly used instruments for assessing either visual difficulty or vision-specific, health-related quality of life have only one to three items on low-luminance tasks that are focused exclusively on night driving (e.g., National Eye Institute Visual Function Questionnaire-25 ³ ; VF-14 ¹¹ ; and the Activities of Daily Vision Scale ¹² ). Although driving is a critical instrumental activity of daily living for many older adults, serving as the primary mode of personal travel, ¹³ many older adults stop night driving or refrain from any driving whatsoever for nonvisual reasons. ^{14 15} Thus, for these individuals, night driving would not be an appropriate way to evaluate low-luminance vision problems, because this is a low-luminance activity they do not engage in for reasons other than vision. Furthermore, night driving is but one of many types of low-luminance activities faced during everyday life. Besides the mentioned questionnaires, other instruments described in the literature either have no nighttime or low-luminance items or have very few (≤2). ^{16 17 18 19}  

Ultimately, widespread public health initiatives for ARM will be targeted at prevention strategies or at arresting early phases of disease before visual acuity impairment is serious, rather than at slowing progression of a severe disease state. Because photoreceptor degeneration is a hallmark of early ARM, ^{5 6} and scotopic dysfunction is an early functional manifestation of the disease, ^{4 8 9 10} it would be advantageous, in evaluating the effects of new treatments targeted at early disease, to be able to measure self-reported visibility problems in reduced light levels. Patient-centered (i.e., self-report) outcome measures have been increasingly viewed as necessary in clinical trials in ophthalmology, serving as secondary endpoints to physiological, anatomic, or psychophysical measures. ^{20 21 22}  

Because currently there is no questionnaire that characterizes visibility problems in low luminance from the patient’s perspective, we sought to develop such a questionnaire as an instrument for use in studies of ARM. Development of a low-luminance questionnaire involved three steps: (1) Identifying the content through focus groups; (2) developing a short version of the questionnaire based on the content analysis and pilot testing of a long version; (3) evaluating psychometric properties of the shortened version. 

Focus Groups and Content Analysis

Methods

Focus group participants were recruited through clinics affiliated with the Department of Ophthalmology, University of Alabama at Birmingham. Participants were contacted by a letter from their ophthalmologist describing the project, followed up by a telephone call inviting the person to participate. In total, nine focus groups were conducted. Six of these consisted entirely of individuals with ARM: two with early ARM, one with intermediate ARM, two with late ARM, and one containing a mixture of all three levels of disease severity. Our strategy was to primarily form groups of persons having similar levels of disease severity to foster perceived commonalities among group members and to promote social facilitation, a common approach in focus group methodology. ^{23 24} For the purposes of focus group recruitment, early ARM was defined as a diagnosis of ARM in the medical record, but no exudative disease or geographic atrophy in either eye, and best corrected acuity in both eyes of 20/70 or better. Intermediate disease was defined as a diagnosis of ARM in the medical record, but no exudative disease or geographic atrophy in either eye, and visual acuity worse than 20/70 in at least one eye. Late ARM was defined as a diagnosis of ARM in the medical record and exudative disease or geographic atrophy in at least one eye. We also included a group of older adults who were free of ARM (did not qualify for ICD-9CM codes for ARM) and had acuity in both eyes of 20/25 or better. Exclusion criteria for all groups were a history of glaucoma, diabetes, optic neuritis, previous ocular trauma, or any other retinal or neurologic condition that would impair vision (other than ARM, as defined for the ARM groups), as determined by a chart review and self-report. 

Because we sought the broadest possible range of comments about low-luminance activities, to facilitate generating questionnaire items, two groups were organized that contained persons with juvenile or early adult-onset retinal degeneration. One group consisted of those with retinitis pigmentosa (ICD9-CM 362.74) and the other those with cone-rod dystrophies or congenital stationary night blindness (ICD9-CM 362.75, 368.61). Although the ultimate questionnaire would not be evaluated on these patients, we viewed their comments as fruitful in identifying ways persons with retinal conditions verbally characterize any low-luminance vision problems that they are experiencing. 

Before each focus group began, informed consent was obtained from participants after the nature and possible consequences of the study were explained. This research was conducted in compliance with the Declaration of Helsinki. The discussion followed a structured protocol modeled after those used in our previous work. ^{25 26} It was led by a trained and experienced facilitator, was audiotaped, and lasted approximately 1 hour. The facilitator stated ground rules for the discussion (e.g., “all opinions valued” and “okay to talk to each other as well as to me”). As an “ice-breaker” participants were asked to introduce themselves by first name and share, in a few words, interests or hobbies. The facilitator then explained that the focus of the discussion was on “vision at night and under low lighting including how well you see under those conditions and how you feel about vision under low lighting.” The protocol began with general, open-ended questions addressing topics such as getting around inside and outside the home, driving, daily visual tasks, reading, writing, adjustment to lighting changes, symptoms, occupational and recreational activities, hobbies, and general well-being. For any problem mentioned, follow-up probes addressed how frequent, important, or serious a problem was; whether the subject had to stop or minimize engaging in an activity; whether tasks took longer; and about feelings of fear, dependency, or frustration. The focus group ended with an inquiry as to whether there were any subjects relevant to the topic that had not been previously addressed and that participants wanted to discuss. 

The audiotapes were transcribed word for word. Comments were identified and coded following a standard multistep content analysis protocol modeled after methods we used previously ^{25 26 27} and as discussed by Holsti. ²⁸ The primary coder read through each transcript to get a general impression of the type of comments made by participants. Next, all comments were highlighted and then, using a set of coding rules, placed into mutually exclusive exhaustive categories according to the main theme of the comment (see the Results section for these categories). During a second pass of the data, these coded comments were categorized as positive, negative, or neutral. A comment was coded as positive if it conveyed that an activity could be done adequately, if pleasant comments were made about it, or if there was an expression of positive affect. Neutral comments were coded when no affective tone could be detected in the comment or when more factual statements were made that were unrelated to disability. Comments were coded as negative or reflecting a “problem” if the respondent reported being limited or having difficulty with activities related to vision, reported a reduced ability to see under certain conditions, or discussed any negative emotion associated with vision. Interrater reliability was assessed by a second independent coder. The second coder read through 15% of the comments and then, using the same coding rules as the primary coder, placed them into categories followed by whether they were positive, negative, or neutral. Interrater reliability was 93% (Cohen’s κ). 

Results

Eighty persons participated in nine focus groups ranging in size from 7 to 13 persons. Table 1shows demographic and visual acuity characteristics of the sample, stratified by diagnostic category. Focus group participants covered a broad range of visual acuity, with better eye acuity ranging from 0.12 logMAR (logarithm of the minimum angle of resolution) (Snellen equivalent, 20/15) to 2.7 logMAR (20/10,000). 

There were 616 comments coded into the following 12 mutually exclusive categories: driving, general vision, psychosocial (e.g., feelings, dependency on others, social relationships and events), mobility, coping behavior, reading, devices (e.g., sunglasses, magnifiers, tinted lenses), adjusting to lighting changes, other activities, color, glare, writing, and peripheral vision (Table 2) . Driving was by far the most frequently mentioned topic. For most categories, the preponderance of comments was negative (i.e., comments identified problems), with 52% of all comments negative, 28% positive, and 20% neutral. The percentage of problem comments falling into various content categories is shown in Figure 1 , with the top six problem areas being driving, psychosocial issues, general vision, adjusting to lighting changes (adaptation), mobility, and reading. Table 3provides information on the distribution of problem comments stratified by visual acuity level, using the mean of the distribution as the cutoff point for forming two acuity categories. This analysis is presented for both the better and worse eyes. In terms of the better eye, when the frequency of problem comments were compared for persons with acuity better than 20/60 with those who had acuity of 20/60 or worse, each problem category was present in both acuity categories with similar proportions. This was also true of the worse-eye analysis, except for the psychosocial category in which more problem comments were made by those with better acuity. The percentage of persons who made problem comments in the top six categories was also not different for those with “good” versus “poor” acuity, and this was the case for both the better- and worse-eye analyses. 

Discussion

The focus group process identified the problem content areas relevant for the development of a questionnaire on vision under low luminance and at night for use with persons having ARM. The most frequently cited problems were driving, psychosocial issues, general vision, adjusting to lighting changes, mobility, and reading. Driving was by far the most frequently articulated problem, and thus it is understandable that nighttime items from previous questionnaires focused on driving. However, it is important to note that low-luminance problems other than driving comprised more than half of the negative focus group comments. In addition, psychosocial comments were a common thread running through the group discussion (as in other focus group studies on ARM ^{25 29} ) suggesting that not only should a low-luminance, night-vision questionnaire address activity difficulties, but also the subject’s feelings relevant to these situations. The problem topics identified were just as likely to be raised by persons regardless of their severity of visual acuity impairment implying widespread generalizability to persons with ARM. One exception is that there were more psychosocial problem comments made by persons with better acuity. At first, this seems paradoxical; however, persons with earlier disease (better acuity) may have more fears or concerns for the future, given the unknown course of their condition; in addition, they are more likely to be closer in time to the date of diagnosis and thus may have had less time to adjust to knowledge that they have a potentially blinding condition. 

Developing a Questionnaire

Methods

Questionnaire items were generated to incorporate the problem categories emerging from the content analysis (Fig. 1) . Ultimately a relatively short questionnaire was desired (<35 items); however, at this stage, we generated more items than eventually were needed, to develop a long version to be used in pilot testing. This preliminary step would generate data informing which items to retain based on high internal consistency. The initial questionnaire had 64 items (examples are in Table 4 ). Response scales for items were based on a five-point scale, with the additional option of the item that did not apply to the person. Some items used a difficulty response scale whose options and corresponding scores (in parentheses) were: no difficulty at all (1), a little difficulty (2), some difficulty (3), a lot of difficulty (4), completely blind under these conditions (5), stopped doing this activity because of my vision (5), and stopped doing this for reasons other than vision or never have done this activity (item not used in calculation of subscale score). Other items had frequency response scales: none of the time (1), a little of the time (2), some of the time (3), most of the time (4), completely blind under these conditions (5), stopped doing this because of vision (5), and stopped doing this for other reasons or never have done this activity (item not used in calculation of subscale score). Subscale scores were computed by scaling individual items from 0 to 100, where 100 represents the highest functional level and 0 the lowest, and then averaging the individual items. 

The 64-item draft version was administered by telephone to focus group participants approximately 6 months after the focus groups and to an additional group of individuals who had ARM but who were unable to attend the focus groups yet who were willing to participate in a telephone survey. 

Factor analysis was performed on the questionnaire responses. Two separate analyses were performed: one on all items except the driving items and the other on the driving items only. There were two analyses, because when all the items were included in a factor analysis, the strength of the relationship among the driving items and with other items obscured the other factors. These analyses defined seven factors, with multiple items in each, that would ultimately serve as the questionnaire’s subscales. In the interest of generating as short a questionnaire as possible, we implemented the following item-reduction plan. Items were excluded that correlated highly with other items in the factor or had severely skewed response distributions. We sought to retain items that addressed topics very frequently mentioned in focus groups, items having approximately normal response distributions (thus more likely to reflect change longitudinally), and items explaining the greatest portion of variance within each defined factor. 

Results

The questionnaire was administered to 132 persons including 72 of the 80 focus group participants. Of the persons from the focus groups who were unavailable for the telephone survey, three were deceased, two withdrew from the study because of family illness, one declined to be interviewed, and two could not be contacted. Table 5shows the demographic and visual acuity characteristics of the telephone questionnaire respondents stratified by diagnostic category. Participants covered a very broad range of visual acuity, with better-eye acuity ranging from −0.12 logMAR (Snellen equivalent, 20/15) to 2.0 logMAR (20/2000). 

Factor analysis revealed seven factors, as listed here with the corresponding Cronbach α coefficient: driving (0.96), extreme lighting (0.92), mobility (0.86), emotional distress (0.84), general dim lighting (0.87), peripheral vision (0.89), and bright sunlight (0.45). All factors had good internal consistency (Cronbach α ≥ 0.8) except for bright sunlight, which was thus dropped from further consideration as a subscale. After the item-reduction strategy implemented as described in the Methods section, we arrived at a 32-item instrument with six subscales (with Cronbach α coefficients): driving (0.94), extreme lighting (0.90), mobility (0.82), emotional distress (0.84), general dim lighting (0.87), and peripheral vision (0.89). These items are listed in Table 4 . 

Discussion

The telephone survey provided information facilitating the development of a shortened 32-item LLQ with six subscales that each had good internal consistency reliability for older respondents with AMD or normal retinal aging. The 32-item questionnaire and scoring instructions, as well as the original longer (64-item) version, are available at www.eyes.uab.edu/tools (provided by the University of Alabama at Birmingham). 

Psychometric Properties of the Shortened, 32-Item Questionnaire

The psychometric properties of the 32-item LLQ were evaluated for older adults who either had age-related maculopathy or normal retinal aging, both of which are known to impact low-luminance vision. ^{4 8 9 10} The extent of potential ceiling and floor effects was characterized. Construct validity was evaluated by examining the association between subscale scores on the 32-item version to the subscale scores of the NEI VFQ-25, ³ a widely used vision-targeted, health-related quality-of-life questionnaire addressing several types of visual task difficulties as well as vision-specific psychosocial issues. Subscale scores for various levels of disease severity were also compared. Criterion validity was evaluated by examining the association between low-luminance questionnaire subscales with psychophysical measures of dark adaptation, including both cone- and rod-mediated parameters. Test–retest reliability was assessed by two questionnaire administrations to the same sample approximately 1 month apart. 

Methods

For all persons participating in establishing the psychometric properties of the 32-item LLQ, demographic information and habitual (“walk-in”) distance visual acuity for each eye using the ETDRS charts ³⁰ were obtained. These samples (characterized below) consisted of older adults exhibiting normal retinal aging or ARM as determined by the masked grading of stereofundus photographs using the Age-Related Eye Disease Study (AREDS) grading system. ³¹ Normal retinal aging was defined as AREDS grade 1; early ARM, grades 2 to 6; intermediate ARM, grades 7 to 9; and advanced ARM, grades 10 to 11. Participants were recruited from the same sources as for the focus groups, but did not include any focus group participants, nor those who were respondents to the previous telephone-administered 64-item LLQ. Exclusion criteria were diagnoses of glaucoma, optic neuropathy, or any ocular conditions other than ARM; refractive error (spherical equivalent) with an absolute value of more than 6 D; neurologic diseases such as Alzheimer’s of Parkinson’s disease; and history of stroke, diabetes serious frailty, or medical conditions expected to lead to mortality or disability within 12 months. Questionnaires were administered in in-person interviews by personnel experienced in administering questionnaires to older adults. 

Construct validity was evaluated by examining the Spearman correlations among the subscales of the LLQ and the NEI VFQ-25. To evaluate criterion validity, LLQ subscale scores for the same sample were evaluated in terms of association with parameters of dark adaptation, as measured psychophysically using procedures described in detail elsewhere (Owsley C et al., manuscript submitted). ^{9 32} Two parameters of dark adaptation were mediated by cone photoreceptor function (cone time constant, cone threshold), two parameters were mediated by rod photoreceptor function (rod slope, rod threshold), and one parameter was influenced by both photoreceptor systems (rod–cone break). Test–retest reliability was evaluated by administering the 32-item questionnaire to a separate sample. The 32-item questionnaire was administered in person by a trained interviewer and then again in person approximately 1 month later. The association between subscale scores at the first and second administration of the questionnaire was computed to evaluate test–retest reliability. 

Results

Construct Validity.

The sample consisted of 125 older adults with mean age 71.5 years (8.7), 58% female and 42% male, and was predominantly white of non-Hispanic origin (96% white, 3% African American, 1% Hispanic). There were 41 (33%) participants who exhibited normal retinal aging, 45 (36%) with early ARM, 19 (15%) with intermediate ARM, 16 (13%) with advanced ARM, and 5 (4%) with ARM of undetermined severity, because fundus grades were unavailable (4%). Visual acuity in the better eye in the sample averaged 0.16 logMAR and ranged from −0.14 logMAR (Snellen equivalent, 20/15) to 0.86 (Snellen equivalent, 20/150). 

The mean and SD for each subscale on the sample as a whole, stratified by disease severity, is in Table 6 , along with the percentage of participants exhibiting the minimum (0) and the maximum (100) scores. Relatively few respondents exhibited a floor effect on any subscales. The subscales were variable with respect to ceiling effects; driving and extreme lighting subscales had minimal respondents with scores of 100, the subscales for mobility, emotional distress, and general dim lighting had approximately one third of respondents at ceiling, whereas nearly half the participants had scores of 100 on the peripheral vision subscale. Subscale scores tended to decrease with increasing disease severity, as defined by fundus photographs and the AREDS grading system (age-adjusted P for trend < 0.003 for all subscales except peripheral vision). 

Table 7lists the age-adjusted correlation coefficients between the LLQ subscales and the NEI VFQ-25 subscales. With very few exceptions, the LLQ subscales exhibited moderate positive correlations with the VFQ subscales. The LLQ driving subscale had its highest correlation to the VFQ driving subscale. The highest correlation on the LLQ emotional distress subscale was with the vision-specific mental health subscale of the VFQ. The LLQ peripheral vision subscale had its highest correlation to the VFQ’s peripheral vision subscale. 

Criterion Validity.

The same sample as used for construct validity was used to evaluate criterion validity. Table 8displays the correlation coefficients between the dark adaptation parameters and LLQ subscale scores. All subscales were significantly associated with aspects of dark adaptation impairment that rely on rod-mediated function. Patients who reported greater difficulties with or emotional distress associated with low-luminance activities were more likely to have a delayed rod–cone break (the time point at which the rod photoreceptors take over after photopigment bleaching), a higher rod threshold, and a flatter rod slope. No subscales were associated with cone-mediated parameters of dark adaptation. 

Test–Retest Reliability.

The sample consisted of 60 older adults with mean age of 71.2 ± 8.1 years, 62% female and 38% male. Fifty-eight of 60 participants were white of non-Hispanic origin; of the remaining two, one was African American and one was Hispanic. There were 19 (32%) participants who exhibited normal retinal aging, 20 (33%) with early ARM, 13 (22%) with intermediate ARM, 7 (12%) with advanced ARM, and 1 (2%) with ARM of undetermined severity because of unavailable fundus photographs. Visual acuity in the better eye averaged 0.17 logMAR (0.16) and ranged from −0.06 logMAR (Snellen equivalent, 20/17) to 0.86 logMAR (Snellen equivalent, 20/145). Table 9displays the Pearson correlation coefficients for assessing the test–retest reliability of each subscale. All subscales had a coefficient ≥0.74, except peripheral vision, which was comparatively lower (r = 0.46). All were statistically significant (P ≤ 0.0003). 

Discussion

The LLQ subscales had good construct validity, in that they were moderately related to NEI VFQ-25 subscales, yet these associations were not so high as to indicate that the two sets of subscales have a high level of redundancy. It is not surprising that the two instruments are related, in that visual task difficulty and its impact on well-being are addressed by the content of both instruments. However, the LLQ has a sole focus on low-luminance and night-time activities, whereas the NEI VFQ-25 does not and instead broadly addresses visual tasks in general regardless of ambient illumination. The LLQ subscales do not display floor effects for persons with ARM and normal retinal aging. A ceiling effect for a substantial number of respondents existed for the peripheral vision subscale. Given that ARM is a condition of central vision where peripheral vision is largely spared, this finding is not surprising. Although the peripheral vision subscale may have limited application to the ARM population, this subscale may be more germane for the study of other types of retinal degenerations more likely to impact the periphery, such as RP, a matter for further study. 

LLQ subscale scores were lower in those with ARM than in those exhibiting normal retinal aging, and among those with ARM, lower scores were associated with greater disease severity. These self-report data mirror the course of ARM pathogenesis in that rod dysfunction and degeneration become progressively worse during disease progression. ^{5 6 7 8 9} What was particularly interesting in the present study is that self-reports of low-luminance task problems on the LLQ showed rod–photoreceptor functional specificity in that LLQ subscale scores were related to rod-mediated dark adaptation parameters but not to cone-mediated parameters. 

Test–retest reliability was acceptably high for all LLQ subscales except for the peripheral-vision subscale where it was less than 0.50. The peripheral vision subscale contained the fewest items of all subscales, was vulnerable to ceiling effects as mentioned earlier, and covered an aspect of visual function that is not typically impaired in ARM. If the peripheral vision subscale and its items are to be deployed in future applications of the LLQ in ARM, this subscale must be developed and evaluated further, to improve its psychometric properties. 

General Discussion

We have developed a questionnaire instrument, the LLQ, for the assessment of problems in low-luminance and nighttime activities in persons with ARM. There are no previously developed questionnaires that fill this gap. For older adults, particularly those with ARM even in its early stages, problems seeing in dim lighting and at night are common visual symptoms that can be verified psychophysically. ^{1 4 8 9 10} These symptoms have a biological basis, in that rod photoreceptors are selectively vulnerable to dysfunction and degeneration in the early phases of ARM. ^{5 6 7} Thus, the LLQ could be useful as a patient-centered outcome for characterizing the daily manifestations of ARM in studying the natural history of the disease and for ultimately evaluating interventions, particularly those targeted at preventing disease emergence and/or arresting early disease progression. 

The LLQ’s psychometric properties support the appropriateness of its future use. Items from the LLQ have real-world validity, because they were generated based on comments of persons with ARM themselves and were not based on the presumptions of “expert opinion.” The subscales have good internal consistency. Construct and criterion validity for ARM have been demonstrated for all subscales, in that scores are lower as a function of increased disease severity are correlated with subscales of a more generally formulated vision-specific, health-related quality of life questionnaire and are associated with rod-mediated parameters of dark adaptation. Test–retest reliability is good for five of six subscales, the exception being the peripheral vision subscale that also exhibited a ceiling effect for a substantial portion of the sample. Challenges faced by the peripheral vision subscale may be related to ARM’s sparing of peripheral vision, although further work will probe these issues in greater depth. Because nearly all the participants during questionnaire development were white, the extent to which the LLQ is appropriate for use with other racial or ethnic groups remains to be determined. 

As effective interventions for preventing early ARM and its progression are developed, the responsiveness of the LLQ subscales to intervention should be evaluated. In addition, the usefulness of the LLQ in studying other retinal degenerations, particularly those that target rod photoreceptors, should be explored. 

 Supported by National Institute on Aging Grant R01-AG04212, National Eye Institute Grant R21-EY14071, the EyeSight Foundation of Alabama, and Research to Prevent Blindness, Inc. CO is a Research to Prevent Blindness Senior Scientific Investigator.

 Submitted for publication September 14, 2005; revised October 21 and November 1, 2005; accepted December 19, 2005.

 Disclosure: C. Owsley, (P); G. McGwin, None; K. Scilley, None; K. Kallies, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Demographic and Visual Acuity Characteristics of Focus Group Participants Stratified by Diagnosis

	ARM					Early-Onset Macular Degeneration	Retinitis Pigmentosa	Elderly Normal Subjects	All Diagnoses Combined
		Early ARM	Intermediate ARM	Late ARM
n	16	10	29	7	9	9	80
Age, Mean (SD)	76.9 (6.5)	78.4 (4.9)	78.9 (6.6)	59.3 (9.9)	57.9 (21.8)	72.7 (3.2)	73.7 (12.0)
Gender, % (n)
Female	62.5 (10)	50 (5)	51.7 (15)	71.4 (5)	77.8 (7)	66.7 (6)	60 (48)
Male	37.5 (6)	50 (5)	48.3 (14)	28.6 (2)	22.2 (2)	33.3 (3)	40 (32)
Race, % (n)
White	100 (16)	100 (10)	96.5 (28)	57.1 (4)	88.9 (8)	88.9 (8)	92.5 (74)
African American	0 (0)	0 (0)	3.5 (1)	42.9 (3)	11.11 (1)	11.11 (1)	7.5 (6)
Visual acuity, Mean logMAR (SD)
Better eye	0.19 (.15)	0.54 (.41)	0.51 (0.38)	0.33 (0.48)	1.28 (1.14)	−0.02 (0.05)	0.46 (0.59)
Worse eye	0.39 (.17)	1.18 (.70)	1.53 (.60)	1.07 (1.10)	1.32 (1.10)	0.04 (0.09)	1.03 (0.84)

Table 2.

 

Frequency of Positive, Negative (Problem), and Neutral Comments within Each Content Category

Content Category	Comment Type(n)					Total Comments
		Positive	Negative (Problem)	Neutral
Driving	47	124	10	181
General vision	35	48	13	96
Psychosocial	7	67	18	92
Mobility	22	19	10	51
Coping behavior	22	2	27	51
Reading	9	18	13	40
Devices	9	2	21	32
Adjusting to lighting changes	1	20	0	21
Other activities	15	4	2	21
Color	5	5	2	12
Glare	2	8	1	11
Writing	1	0	3	4
Peripheral vision	0	4	0	4

  n = 616 comments.

Figure 1.

 

Number of problem comments made in the focus groups stratified by topic of comment.

Table 3.

 

Distribution of Focus Group Problem Comments by Acuity

Problem Comment Categories*	Better Eye					Worse Eye
		Better than 20/60 n (%)	20/60 or Worse n (%)	P , †	Better than 20/200 n (%)	20/200 or Worse n (%)	P , †
Driving	80 (36.9)	44 (42.3)	0.35	73 (36.0)	51 (43.2)	0.20
Psychosocial	51 (23.5)	16 (15.4)	0.09	53 (26.1)	14 (11.9)	0.003
General vision	29 (13.4)	19 (18.3)	0.25	25 (12.3)	23 (19.5)	0.08
Adjusting to lighting changes	17 (7.8)	3 (2.9)	0.08	16 (7.9)	4 (3.4)	0.11
Mobility	13 (6.0)	6 (5.8)	0.94	13 (6.4)	6 (5.1)	0.63
Reading	12 (5.5)	6 (5.8)	0.93	10 (4.9)	8 (6.8)	0.49

*  Categories with fewer than 10 problem comments are not presented.

†  By χ² test.

Table 4.

 

Items in the 32-Item Version of the LLQ

No.	Item	Response Scale
1	Do you have difficulty seeing in bright sunlight?	Difficulty
2	Do you have difficulty seeing in fluorescent lighting, like that found in stores and offices?	Difficulty
3	Do you have difficulty seeing people’s faces in a hallway when direct sunlight is behind them?	Difficulty
4	Do you have difficulty reading menus in dimly lit restaurants?	Difficulty
5	Do you have difficulty reading the newspaper without good lighting?	Difficulty
6	Do you get upset because you have difficulty seeing while driving in the rain?	Frequency
7	Do you have difficulty reading material printed on dark-colored paper?	Difficulty
8	Do you have difficulty seeing dark-colored cars while driving at night?	Difficulty
9	Because of your vision, are you bothered that you have difficulty moving around in a darkened theater?	Frequency
10	Because of your vision, do you have difficulty going out to nighttime social events, such as sporting events, the theater, friends’ homes, church, or restaurants?	Difficulty
11	Do you depend on others to help you because of your vision at night or under poor lighting?	Frequency
12	Do you worry or are you concerned that you might fall at night because of your vision?	Frequency
13	Do you have difficulty seeing colors at night?	Difficulty
14	Do you have difficulty seeing furniture in dimly lit rooms with dark floors?	Difficulty
15	Do you have difficulty seeing at night?	Frequency
16	Do you have difficulty seeing in poor lighting conditions, such as dusk or dawn or in a poorly lit room?	Frequency
17	Do you have difficulty with depth perception at night?	Difficulty
18	Do you have difficulty seeing in candlelight?	Difficulty
19	Do you have difficulty seeing when you visit other people’s homes because there is not enough light?	Difficulty
20	Do you have difficulty seeing under kitchen counters or in cabinets or closets because there is not enough light?	Difficulty
21	Do you have difficulty with your peripheral vision under poor lighting conditions?	Difficulty
22	Do you have difficulty with your peripheral vision at night?	Difficulty
23	Do you have difficulty with your peripheral vision in bright sunlight?	Difficulty
24	Do you have difficulty reading street signs when driving at night?	Difficulty
25	While driving at night, do you have difficulty with headlights from oncoming cars?	Difficulty
26	Have you limited driving in the rain because of difficulty seeing?	Frequency
27	Do you limit your driving at night because of your vision?	Frequency
28	Do you have difficulty seeing while driving at dawn or dusk because of glare?	Difficulty
29	Do you worry or are you more concerned that you may make a mistake at a social event because you can’t see well enough under poor lighting conditions? (for example, getting food on a fork, recognizing people, or reading the menu in a dimly lit restaurant)	Frequency
30	Do you feel bad or depressed about your ability to see at night or under poor lighting conditions?	Frequency
29	While driving at night, do you have difficulty judging the distance between you and other moving cars?	Difficulty
30	While driving at night, do you have difficulty judging the distance to your turn-off or exit?	Difficulty
31	Do you feel bad or depressed because your vision at night or under poor lighting keeps you from doing all that you would like to do?	Frequency
32	Do you feel bad or depressed that you aren’t able to help others as much as you want because of your vision at night or under poor lighting?	Frequency

 Items in each subscale: driving (24, 25, 26, 27, 28), extreme lighting (1, 2, 3, 4, 5, 7, 8), mobility (9, 10, 11, 12, 13, 14), emotional distress (29, 30, 31, 32), general dim lighting (15, 16, 17, 18, 19, 20) and peripheral vision (21, 22, 23).

Table 5.

 

Demographic and Visual Acuity Characteristics of Telephone Questionnaire Participants Stratified by Diagnosis

	ARM					Early-Onset Macular Degeneration	Retinitis Pigmentosa	Elderly Normal	All Diagnoses Combined
		Early ARM	Intermediate ARM	Late ARM
n	32	14	39	13	10	24	132
Age, mean (SD)	76.7 (10.2)	81.6 (5.8)	81.4 (5.6)	58.5 (13.1)	60.4 (19.7)	74.0 (4.7)	75.1 (12.0)
Gender, % (n)
Female	71.9 (23)	64.3 (9)	60.0 (23)	76.9 (10)	70.0 (7)	62.5 (15)	65.9 (87)
Male	28.1 (9)	35.7 (5)	41.0 (16)	23.1 (3)	30.0 (3)	37.5 (9)	34.1 (45)
Race, % (n)
White	96.9 (31)	100 (14)	100 (39)	84.6 (11)	80 (8)	70.8 (17)	93.9 (124)
African American	3.1 (1)	0 (0)	0 (0)	15.4 (2)	10 (1)	20.8 (5)	3.8 (5)
Unknown	0 (0)	0 (0)	0 (0)	0 (0)	10 (1)	8.3 (2)	2.3 (3)
Visual acuity, Mean logMAR (SD)
Better eye	0.17 (0.25)	0.55 (0.37)	0.62 (0.48)	0.43 (0.50)	0.68 (0.65)	0.03 (0.13)	0.39 (0.46)
Worse eye	0.32 (0.27)	0.93 (0.21)	1.13 (0.43)	0.75 (0.85)	0.77 (0.57)	0.12 (0.26)	0.67 (0.57)

Table 6.

 

Mean, Standard Deviation, and Floor and Ceiling Effects for LLQ Subscales for a Sample of Persons with ARM or Normal Retinal Aging

Subscale	Total Sample			Normal Aging (n = 41)			Early ARM (n = 45)			Intermediate ARM (n = 18)			Advanced ARM (n = 16)			Floor Effect (%)	Ceiling Effect (%)
		M	SD	M	SD	M	SD	M	SD	M	SD
Driving	62.5	31.9	75.6	26.5	62.0	28.9	43.5	34.9	55.0	36.2	12.8	9.6
Extreme lighting conditions	70.7	18.8	80.1	17.0	69.8	14.8	61.6	18.9	59.1	23.3	0	5.6
Mobility	86.4	15.5	92.2	10.4	86.1	16.2	80.6	14.7	79.0	18.6	0	26.4
Emotional distress	87.9	17.7	94.2	10.4	88.9	16.0	84.0	16.4	76.2	20.8	0.8	37.6
General Dim-lighting problems	83.2	16.8	90.3	11.7	81.6	16.2	78.1	17.2	74.0	21.8	0	21.6
Peripheral vision	84.7	18.7	90.4	14.1	83.7	18.8	75.9	18.9	84.4	24.1	0	45.6

  n = 125; 5 persons whose ARM severity could not be confirmed by fundus photos were not included in the subgroup analysis. M, mean.

Table 7.

 

Age-Adjusted Pearson Correlations between LLQ and NEI VFQ-25 Subscales

NEI VFQ-25 Subscales	LLQ Subscale
		Driving	Extreme Lighting	Mobility	Emotional Distress	General Dim Lighting	Peripheral Vision
General health	0.15	0.03	0.15	0.19	0.15	0.09
General vision	0.56*	0.72*	0.44*	0.56*	0.56*	0.38*
Near vision	0.67*	0.79*	0.49*	0.60*	0.63*	0.46*
Distance vision	0.70*	0.79*	0.70*	0.71*	0.69*	0.53*
Driving	0.86*	0.80*	0.52*	0.47*	0.54*	0.43*
Peripheral vision	0.43*	0.47*	0.57*	0.54*	0.50*	0.65*
Color vision	0.45*	0.49*	0.55*	0.37*	0.44*	0.36*
Ocular pain	0.27*	0.30*	0.40*	0.35*	0.31*	0.19*
Vision-specific role difficulties	0.65*	0.71*	0.63*	0.63*	0.59*	0.56*
Dependency	0.60*	0.71*	0.63*	0.67*	0.51*	0.55*
Social functioning	0.59*	0.71*	0.60*	0.59*	0.58*	0.52*
Mental health	0.55*	0.69*	0.60*	0.80*	0.57*	0.41*

  n = 111 (of 125 enrolled) because 14 participants did not have complete data.

*  Statistically significant (P ≤ 0.05, two-tailed).

Table 8.

 

Pearson Correlation Coefficients between LLQ Subscales and Dark Adaptation Parameters

Parameters of Dark Adaptation	LLQ Subscales
		Driving	Extreme Lighting	Mobility	Emotional Distress	General Dim Lighting	Peripheral Vision
Cone time constant	.00	−.01	−.02	−.02	.00	−.06
Cone threshold	.06	.10	−.01	.14	.04	.10
Rod cone break	−.39*	−.36*	−.23*	−.28*	−.26*	−.16
Rod slope	.40*	.34*	.28*	.25*	.27*	.19*
Rod threshold	.43*	.38*	.33*	.37*	.30*	.25*

  N = 125.

*  Statistically significant (P ≤ 0.05, two-tailed).

Table 9.

 

Test–Retest Reliability over a 1-Month Interval for the LLQ Subscales

Subscale	Pearson Correlation Coefficient
Driving	0.88
Extreme lighting	0.75
Mobility	0.74
Emotional distress	0.78
General dim lighting	0.82
Peripheral vision	0.46