OBJECTIVE
To determine if home environmental features (ie, lighting and home hazards) are associated with difficulties in instrumental activities of daily living (IADLs) in persons with glaucoma.
DESIGN
Cross-sectional study.
SETTINGS AND PARTICIPANTS
A total of 174 adults with suspect or primary glaucoma were recruited.
METHODS
IADLs, fear of falling (FoF), and glaucoma quality of life (GQL) were assessed via questionnaire. FoF and GQL were prespecified as negative controls to assess discriminant/external validity, as they are not expected to be influenced specifically by in-home lighting or hazards. We classified IADL difficulty as a binary outcome (≥1 IADL difficulty vs none). Person-measure scores for FoF and GQL were calculated in logits using Rasch modeling. The home environment assessment for the visually impaired (HEAVI) tool assessed homes for the total number of hazards, frequency of hazards among graded items, and average home lighting. Multivariable logistic and linear regression models evaluated associations between home environmental measures with IADL difficulty and FoF and GQL scores, controlling for severity of visual field damage, age, race, sex, comorbidity, and polypharmacy.
MAIN OUTCOME MEASURES
Association between home environmental features with difficulties in IADLs, FoF, and GQL.
Results
Better home lighting was associated with less difficulty completing IADL tasks (odds ratio [OR] = 0.88 per 0.1 log unit light increment, 95% CI = 0.77-1.00, P =.04). No association was found between the number of home hazards (OR = 0.97 per 10 additional hazard, 95% CI = 0.62-1.53, P =.89) or the frequency of home hazards among graded items (OR = 1.10 per 10% increment in hazard frequency, 95% CI = 0.65 to 1.86, P =.73) with IADL difficulty. No significant associations were noted between lighting levels, frequency or number of home hazards with FoF or GQL scores.
Conclusion
Better home lighting was associated with less frequent difficulty with IADLs (but not FoF or GQL) in a cohort enriched for glaucoma; no associations were noted with any measure for home hazards. Lighting may be able to improve specific aspects of function in persons with glaucoma.
I nstrumental activities of daily living (IADLs) represent high-level activities that entail interacting with the environment and using cognitive functions to achieve tasks including managing finances, medication, preparing meals, taking care of transportation and mobility, shopping for essentials, and managing the home (eg, cleaning and doing laundry). , As people age, sensory limitations, physical limitations, and cognitive decline challenge independent performance in managing IADLs. ,, Hence, IADLs are frequently used by health care practitioners and caregivers to assess the functional state of older adults and determine whether they need support or care. IADL limitations may co-occur with, or result in, difficulties with other aspects of well-being including quality of life. ,,,, One factor influencing IADL ability and other patient-reported measures of well-being is poor vision, , with specific diseases such as glaucoma and AMD associated with a higher likelihood of IADL difficulty. A large body of literature provides plausible mechanisms linking glaucoma to IADL difficulty, with severity of visual field damage associated with less mobility, poor levels of activity, , an increased risk of falls, , and difficulty in reading. ,
Unfortunately, for glaucoma and other diseases that have no therapy to reverse visual damage, IADL disability cannot be remedied by routine ophthalmic care. ,,, An alternative approach for addressing IADL disability in conditions such as glaucoma is to address the environment in which IADLs are completed, especially as the large majority of IADLs are performed in the home. Indeed, home environment features such as lighting have been associated with fewer falls, more in-home activity, and gait features indicative of greater confidence. Moreover, prior studies have demonstrated that, in a group of patients with glaucoma, those with greater disease severity (at more risk of disability) have levels of lighting and home hazard levels similar to those with less severe disease (at lower risk of disability), suggesting that individuals with more severe vision loss do not alter their homes for safety. However, prior studies have not evaluated whether home features (eg, lighting and home hazards) affect the ability of those with visual impairment to perform IADLs.
As part of a prospective study involving an in-home assessment to judge lighting and in-home hazards, we used a validated home assessment tool to identify whether home environmental features (ie, lighting and/or home hazards) were associated with difficulties in IADLs in individuals with a range of glaucoma damage. We hypothesized that poor lighting within the home, and more home hazards, would be associated with greater IADL difficulty. We also explored whether the severity of VF damage interacted with home environmental features (ie, lighting and hazards) to produce greater difficulty with IADL difficulty in glaucoma. Finally, we also explored whether home environmental features (ie, lighting and hazards) would also show associations with fear of FoF and GQL. Given that IADLs largely focus on activities within the home, whereas FoF and GQL affect activities both within and outside the home, we hypothesized that home environmental features would show associations with IADL difficulty but weaker or absent associations with FoF and GQL (negative controls).
METHODS
All participants provided written informed permission for study procedures, and the Johns Hopkins School of Medicine Institutional Review Board approved the study protocol. The study adhered to the Declaration of Helsinki’s principles. The study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
STUDY PARTICIPANTS
From September 2013 to March 2015, study participants were recruited from the Johns Hopkins Wilmer Eye Institute glaucoma clinic, into the Falls in Glaucoma Study (FIGS). Participants were eligible for the study if they turned 60 years of age or older by the conclusion of the 3-year study period, had a primary open-angle, primary angle closure, pigmentary, or pseudoexfoliation glaucoma diagnosis, or were at high risk of, or suspected of having, glaucoma. Patients who had 20/40 vision or worse in either eye from any ocular condition other than glaucoma, were bedridden or in a wheelchair, had been hospitalized within the previous month, had undergone ocular or nonocular surgery within the previous 2 months, had a history of stroke or any other neurologic condition that damaged the visual field (VF), or lived more than 60 miles from Baltimore were excluded from the study.
VISUAL ASSESSMENT
The Humphrey Visual Field Analyzer (Carl Zeiss Meditec) Standard 24-2 Swedish Interactive Testing Standard Algorithm (SITA-Standard) was used to evaluate severity of glaucoma damage in study participants. , VF testing was conducted independently for each eye, and integrated visual field (IVF) sensitivity was calculated by combining VF data from both eyes. Spatial matching points of the VF between the 2 eyes were combined using the maximal sensitivity approach, and the average sensitivity across the IVF was computed. , Participants were categorized as having mild (>28 dB), moderate (23-28 dB), or severe (<23 dB) VF impairment based on IVF sensitivity, as previously described. A back-lit Early Treatment Diabetic Retinopathy Study (ETDRS) chart was used to separately assess baseline visual acuity (VA) at 4 m in each eye. The MARS chart was used to assess binocular contrast sensitivity (Mars Perceptrix). All visual testing was performed with participants’ presenting correction (if any).
HOME ENVIRONMENTAL ASSESSMENT
Potential home hazards were evaluated using the Home Environment Assessment for the Visually Impaired (HEAVI) tool. , To get the most accurate assessment of the home, participants were requested not to clean their houses or to alter their homes in other ways (eg, improving lighting and decluttering) before the coordinator arrived. In 8 distinct areas of the house—the (1) foyer, (2) living room, (3) dining room, (4) kitchen, (5) bedroom, (6) bathroom, (7) stairs, and (8) hallway—a total of 127 potential risks/hazards were evaluated by a trained grader. Participants were asked to set the light and window coverings to a standard setting matching typical use for that time of day to gauge the illumination level. A digital light meter (Dr. Meter model LX1330B; Hisgadget Inc) was used to measure the amount of light in each room in lux. Lighting measurement was log-transformed and averaged across rooms of the home to generate the average of lighting in the house; the total number of hazards was calculated based on all the hazards found across the room in each home, and a frequency of hazards was calculated using the total number of hazards detected in the home, normalized to the number of possible hazards in the home (based on the number of graded items). Lighting levels were log-transformed, such that each 0.1 unit in log(lux) represents a roughly 25% increment in lighting.
Assessment of IADLs
IADLs were evaluated with previously used questionnaires. , To rate how much difficulty participants had with each IADL, they were given 4 options: “No Difficulty,” “With Difficulty, but without help,” “With Help,” or “Unable to Perform Task.” Grocery shopping, money management, meal preparation, medicine administration, telephone use, heavy and light housework, and transportation to locations beyond walking distance, were the 8 activities assessed. For this analysis, difficulty with IADLs was dichotomized to any reported difficulty with 1 or more IADLs vs no difficulty reported with any IADLs. ,
Evaluation of FoF
FoF was evaluated using the University of Illinois Chicago FoF questionnaire. , FoF was prespecified as a negative control instrument to assess discriminant/external validity. Participants were asked to rate their level of fear in different scenarios using 3 possible answers: “Not worried at all,” “A little/moderately worried,” or “Very worried.” 28 Person-measure scores representing FoF levels were calculated in logits using a Rasch-calibrated model, with higher scores indicate greater FoF.
Assessment of GQL
GQL was assessed using the GQL-15. , Participants were asked to score difficulty with each described task as “none,” “a little bit,” “some,” “quite a lot,” “severe,” or “not applicable. GQL was prespecified as a negative control instrument to assess discriminant/external validity. Person-measure scores were expressed in logits using Rasch modeling, with higher scores indicating worse GQL. ,
EVALUATION OF COVARIATES
Standardized questionnaires were used to gather demographic/health data, including age, gender, race, and medical history. Participants were asked if they had received a diagnosis for any of the 15 illnesses including arthritis, a history of hip fracture, back pain (backache or lumbago), prior heart attack, angina or chest pain, congestive heart failure, peripheral vascular disease, hypertension, diabetes, emphysema, asthma, stroke, Parkinson, nonskin cancer, and vertigo or Meniere disease. The total number of conditions was reported to calculate a comorbidity index. To determine the number of systemic prescription medications used, either direct bottle observation or participant report was used. The use of 5 or more prescribed non–eye drop drugs was classified as polypharmacy. The Jamar Hand Dynamometer (Sammons Preston) was used to test grip strength 3 times with the dominant hand, and a microFET2 Dynamometer (Hoggan Scientific LLC) was used to measure leg strength twice on each leg. The maximum leg and grip strength between the measurements was reported here in kilograms of force.
STATISTICAL ANALYSIS
Multivariable logistic regression models were used to evaluate the association between specific home features (average log home lighting, total number of hazards, frequency of items graded as hazardous) with IADL difficulty. Additional multivariable linear regression models were used to evaluate the association of these home features with either fear of falling or glaucoma quality of life. Models controlled for age, race, sex, comorbidities, polypharmacy, and IVF sensitivity. To define whether the effects of environmental risks on IADLs, FoF, and GQL varied over the spectrum of visual field damage severity, additional models included an interaction between IVF and each of the home hazard metrics. Data were analyzed using Stata Statistical Software, release 15 (StataCorp).
RESULTS
PATIENTS SAMPLE CHARACTERISTICS
For the 174 study participants, average age was 71.0 years (SD 7.6). Nearly 30% (29.9%) of participants identified as Black, with 54.0% male. Just under one-third (32.8%) of participants were employed, 18.9% lived alone, and about 86% had at least some college education or higher. More than half (62.1%) had more than 1 comorbid illness, and 30.5% were taking 5 or more systemic medications. With regard to vision, the median integrated visual field (IVF) sensitivity was 28.1 dB (with 31.0 dB considered normal). Median values for mean deviation (MD) measurements in the better eye and worse eye were–2.5 dB (ranging from–29.63 to 2.59 dB) and–5.7 dB (ranging from–31.89 to 1.55 dB). The better-eye VA measured at a median logMAR value of 0.06. Median binocular log contrast sensitivity was 1.72 ( Table ).
TABLE
Characteristics of Study Participants (N = 174).
| Demographics | Values |
|---|---|
| Age, y, mean (SD) | 71.0 (7.6) |
| Black, n (%) | 52 (29.9) |
| Male sex, n (%) | 94 (54.0) |
| Employed, n (%) | 57 (32.8) |
| Lives alone, n (%) | 33 (18.9) |
| Education, n (%) | |
| Less than high school | 7 (4.0) |
| High school | 18 (10.3) |
| Some college | 22 (12.6) |
| Bachelor’s degree | 45 (25.9) |
| More than bachelor’s degree | 82 (47.1) |
| Health | |
| Comorbid illnesses >1, n (%) | 108 (62.1) |
| Polypharmacy, n (%) | 53 (30.5) |
| Body mass index, mean (SD) | 27.2 (5.0) |
| Grip strength, kg, mean (SD) | 32.0 (10.3) |
| Lower body strength, kg, mean (SD) | 18.0 (6.4) |
| Vision, median (IQR) | |
| IVF sensitivity, a dB | 28.10 (25.90, 29.73) |
| MD: better eye | –2.5 (–5.38,–0.68) |
| MD: worse eye | –5.71 (–12.96,–2.64) |
| Better-eye acuity, logMAR | 0.06 (–0.02, 0.16) b |
| Binocular log CS | 1.72 (1.64, 1.76) |
CS = contrast sensitivity, IVF = integrated visual field, MD = mean deviation.
RELATIONSHIP BETWEEN ENVIRONMENTAL AND VISION MEASURES WITH IADL DIFFICULTY
Lighting levels were associated with difficulty in completing IADLs, with people reporting less difficulties in IADLs with better lighting. Specifically, there was a 12% lower likelihood of having any IADL difficulty when lighting (lux) was 0.1 log units brighter (OR = 0.88 per 0.1 log units increment in light, 95% CI = 0.77-1.00, P =.04; Figure 1 ). No association was seen between either the number of home hazards (OR = 0.97 per 10 more hazards, 95% CI = 0.62-1.53, P =.89) or the frequency of home hazards among graded items (OR = 1.10 per 10% increase in frequency of hazards, 95% CI = 0.65-1.86, P =.73) with IADL difficulty. In models that included home environmental features, IVF sensitivity was not significantly associated with the likelihood of IADL difficulty ( P >.05 for models including home lighting, and number and frequency of home hazards). No interactions were noted between home environmental feature (lighting, total number of hazards, frequency of hazards) and IVF sensitivity with regard to the likelihood of IADL difficulty ( P >.53 for all), suggesting that the relationship between the home environmental features and IADLs does not differ by the level of visual impairment.
