Factors Associated With Visual Impairment and Eye Care Utilization: The International Mobility in Aging Study

Abstract

Objective:To examine factors associated with visual impairment (VI) and eye care in the International Mobility in Aging Study (IMIAS). Method: IMIAS data were analyzed (N = 1,995 with ages 65-74). Outcomes were VI defined as presenting visual acuity worse than 6/18 in the better eye and eye care utilization assessed by annual visits to eye care professionals. The Hurt–Insult–Threaten–Scream (HITS) questionnaire requested information on domestic violence. Results: Among men, VI varied from 24% in Manizales (Colombia) to 0.5% in Kingston (Canada); among women, VI ranged from 20% in Manizales to 1% in Kingston; lifetime exposure to domestic violence was associated with VI (odds ratio [OR] = 1.87; 95% confidence interval [CI] = [1.17, 3.00]). Eye care utilization varied from 72% in Kingston’s men to 25% in Tirana’s men; it was associated with domestic violence (prevalence ratio [PR] = 1.3; 95% CI = [1.1, 1.6]). Discussion: VI is more frequent where eye care utilization is low. Domestic violence may be a risk factor for VI.

Keywords

vision domestic violence aging global health

Introduction

Significant effort has been made to estimate the prevalence of global visual impairment and blindness with the latest estimates being 191 million and 32 million, respectively (Pascolini & Mariotti, 2012; Resnikoff et al., 2004; Stevens et al., 2013). Yet, population-based data are lacking particularly on the aging population, especially in the developing world where the burden of visual impairment is greatest.

The International Mobility in Aging Study (IMIAS), conducted in 2012, can be used to provide these essential data. IMIAS is a longitudinal population-based study of life-course factors and their relationship with mobility and physical health in older adults conducted in Manizales (Colombia), Tirana (Albania), Natal (Brazil), Saint-Hyacinthe (Canada), and Kingston (Canada; Zunzunegui et al., 2015). The IMIAS rationale was to test if the variability in the gender gap in mobility and physical function of older adults could be explained by gender inequality, by comparing older adults who reside in contexts that provide a wide range of exposures, health and functional outcomes. IMIAS includes older adults residing in five cities located in Colombia, Albania, Brazil, and Canada, countries differing in gender inequality. According to the Gender Inequality Index published in 2015, Colombia ranks 92, Albania 45, Brazil 97, and Canada 25 (United Nations Development Programme [UNDP], 2015).

Visual acuity was measured due to epidemiological evidence on the relationship between vision and mobility (Fenwick et al., 2016). Prior peer-reviewed research on the prevalence of visual impairment in adults in Colombia, Albania, Brazil, and Canada has been limited or does not exist (Maberley et al., 2006; Robinson et al., 2013; Salomao et al., 2008; Schellini et al., 2009). In our current research, we have unveiled the relatively high exposure to lifetime physical violence in IMIAS (Zunzunegui et al., 2015), and have examined its interactions with family living arrangements (Guedes et al., 2015) and its associations with physical function (Guedes et al., 2016) and c-reactive protein (CRP) in old age (Sousa et al., 2014). In this article, we advance the hypothesis that lifetime exposure to family physical violence constitutes a risk factor for visual impairment.

The goal of this article was to identify factors associated with visual impairment and eye care utilization. In addition to the well-known demographic (age and sex) and socioeconomic (education and income) risk factors for visual impairment, we examine the associations between visual impairment and family physical violence exposure, taking into account the diverse living arrangements in these very different settings. We also examine factors associated with eye care utilization, including predisposing factors (gender, education, history of family physical violence for potential eye trauma) and enabling factors (income, urban, and living arrangements) that can affect eye care utilization in old age.

Method

Context

Manizales is a city with 400,000 people and lies in the Andes Mountains in Colombia. Tirana is the capital city of Albania and has approximately 700,000 people. Natal is a coastal city in northeast Brazil and has about 800,000 people. Saint-Hyacinthe is a French-speaking city of 53,000 people in Quebec, Canada. Kingston is an English-speaking city of 123,000 people in Ontario, Canada. Colombia has a two-tier system with public and private health care insurances, whereas Albania, Brazil, and Canada have universal health care systems.

Recruitment

In 2012, approximately 200 men and 200 women aged 65 to 74 years and living in the community were recruited through neighborhood primary care center registers. In Manizales, a random sample of older adults registered in the city public health care (which covers 82% of the older adult population) was drawn. In Tirana and Natal, random samples of elderly people registered at their neighborhood health centers were drawn. Response rates at these sites were close to 100%. In Saint-Hyacinthe and Kingston, potential participants received a letter from their family doctor inviting them to contact our field coordinator to set up an appointment for a home visit (Zunzunegui et al., 2015). This indirect method of recruitment in the Canadian cities was required because the ethics committees for the Canadian sites did not authorize researchers to communicate directly with potential participants to invite them to participate in a study. In Saint-Hyacinthe, the sample was stratified by neighborhood, whereas in Kingston this stratification was not possible.

Exclusion Criteria

Participants were excluded if they had four or more errors on the orientation scale of the Leganes Cognitive Test (De Yebenes et al., 2003), administered at the beginning of the interview. Low scores on this test were considered indicative of inability to complete the study procedures. Exclusions ranged from five in Natal to none in Kingston.

Ethical Considerations

The research ethics committees at the Universidad de Caldas (Manizales, Colombia), the Albanian Institute of Public Health (Tirana, Albania), the Universidade Federal do Rio Grande do Norte (Natal, Brazil), the Centre de Recherche du Centre Hospitalier de l’Université de Montréal (Montreal, Canada), and Queen’s University (Kingston, Canada) approved the study. Research complied with the tenets of the Declaration of Helsinki. Written informed consent was obtained from all subjects prior to study participation. Illiterate individuals had the form read to them and signed using a fingerprint.

Data Collection

Data were collected through personal interviews conducted in the participants’ homes with the exception of Manizales, where vision tests were conducted at the local hospital. Questionnaires were administered in Spanish, Albanian, Portuguese, French, and English.

Outcomes

This analysis had two outcomes: visual impairment and eye care utilization. Visual impairment was defined using the World Health Organization definition (presenting visual acuity worse than 6/18 in the better eye). Presenting Visual acuity (habitual) was measured at 2 m using the 2-m Early Treatment Diabetic Retinopathy Study (ETDRS) Tumbling E chart, which is useful in populations with high degrees of illiteracy (Taylor, 1978). Participants were instructed to take the test with their glasses if they normally wore glasses for tasks done at a distance. The tumbling E chart was used because we expected a relatively large proportion of illiteracy in the Latin American cities’ elderly population. The advantage of using presenting visual acuity is that it takes into account visual impairment due to uncorrected refractive errors. Letter-by-letter scoring was used, and the test was stopped when participants answered four of five letters incorrectly on one line. Forced choice procedures were employed in which a participant was encouraged to guess. Scores were converted to logMAR.

Eye care utilization was assessed by questions about whether participants had seen an eye care provider in the last year (yes, no). Wearing glasses to see far away was registered as the answer to the question: “Do you use eyeglasses or contact lenses or both to see far away?” (possible answers: yes or no).

Risk Factors

Physical domestic violence exposure related to trauma

Participants were asked the Hurt–Insult–Threaten–Scream (HITS) domestic violence screening questionnaire (Sherin, Sinacore, Li, Zitter, & Shakil, 1998). In this analysis, we focused on the question whether the participant had ever in his or her lifetime had a family member who had physically hurt him or her. Responses included never, rarely, sometimes, fairly often, or frequently. Those who had been hurt sometimes, fairly often, or frequently were considered to have been exposed to domestic physical violence. Thus, we dichotomized this variable into never or rarely versus sometimes, fairly often, or frequently.

Living arrangements

Living with spouse and children in Latin America and Albania is a proxy for a good social situation, while living alone or with others is associated with destitution (very different from North America where living alone is the norm for unmarried older adults; Zunzunegui et al., 2015; Zunzunegui, Beland, & Otero, 2001). Living arrangements could be associated with visual impairment as a result of social disadvantage. Participants were asked if they were living alone. In the case of a negative answer, they specified if they were living with spouse, children, family, or others. Independent dichotomous variables were constructed for each of these living arrangements.

Demographic and socioeconomic variables

Questions were asked about demographic variables such as age, sex, education, and income. Education was measured by the number of years of schooling. Participants were asked to report their household monthly income before taxes, and were then placed in one of four income categories. Cutoff points were adapted for each site. People were placed in an income category based on its relationship to the country-specific minimum wage (less than the country-specific minimum wage), lower middle (country-specific minimum wage), upper middle (2-3 times the country-specific minimum wage), or high (4 or more times the country-specific minimum wage) income. In the Canadian sites, cutoff points were defined by annual household income of less than Can$20,000, between Can$20,000 and Can$40,000, between Can$40,000 and Can$60,000, and more than Can$60,000. Income is influenced by pension laws in each country. There is no universal pension in Colombia; Albania grants a pension to all citizens above age 65, Brazil grants a pension to all citizens above age 60, and Canada provides a pension to most citizens above age 65 (except the wealthy).

Medical conditions and smoking

Diabetes and smoking have been identified as risk factors for visual impairment (Krishnaiah et al., 2005; Leasher et al., 2016).

Statistical analysis

Means, standard deviations, and percentages were calculated. Differences by site were tested by ANOVA or chi-square test. Differences by outcome status were tested by Student’s t test or chi-square test.

As visual impairment in Canada was rare (3% and less), we did not carry out multivariate analyses of visual impairment in the Canadian cities. For Manizales, Natal, and Tirana, we assessed associations in multiple logistic regression models. Standard errors were adjusted for the clustering within the sites by using clustered sandwich estimators (Rabe-Hesketh & Skrondal, 2008).

As the visits to an eye care professional were relatively common (ranging from 40%-60%) and as we were seeking a direct estimation of the prevalence ratio (PR), we used Poisson’s regression with a robust variance correction (Williams, 2000).

To test for heterogeneity and to obtain pooled estimates of the associations between domestic physical violence and visual impairment and eye care, we carried out meta-analysis; first, to examine the association between domestic physical violence and vision impairment (three non-Canadian sites) and second, between physical violence and eye care (all sites). For these meta-analyses, the adjusted estimates of associations in the site-specific analyses and their standard errors were used.

All analyses were run in Stata Version 11.2 (College Station, Texas).

Results

Characteristics of the Study Samples

The mean age at each site was 69 years, and there were approximately equal numbers of men and women. The participants in Kingston had the most formal education with an average of 16 years, whereas the participants in Natal had the least at 5 years. The majority of people in Kingston were in the high-income category, whereas the majority of people in Manizales were in the low-income category. Physical violence from a family member was reported more often in Manizales and Kingston with approximately 8% of people reporting it sometimes, fairly often, or frequently in their lifetime (Table 1). Frequency of visual impairment in the study samples is shown in Figure 1 with significant differences across cities. Men presented higher frequencies of visual impairment than women in the Canadian cities and lower frequencies than women in the non-Canadian samples; the sex difference was only statistically significant in Tirana (p = .02). Percentages reporting eye care in the last year varied significantly across sites (Figure 2). The proportion wearing glasses was as follows: 70.4% in Kingston, 75.7% in Saint-Hyacinthe, 58.1% in Tirana, 46.2% in Manizales, and 70.7% in Natal.

Table 1.

Sociodemographic, Economic, and Social Characteristics of the Study Samples in IMIAS at Baseline.

	Kingston, Canada (n = 398)	Saint-Hyacinthe, Canada (n = 401)	Tirana, Albania (n = 394)	Manizales, Colombia (n = 400)	Natal, Brazil (n = 402)	p value
Age, mean (SD) years	69.1 (2.7)	68.6 (2.7)	69.4 (3.1)	69.3 (3.0)	69.2 (2.7)	.001
Sex, %
Women	53.3	52.4	52.3	50.5	52.2	.958
Men	46.7	47.6	47.7	49.5	47.8
Education, mean (SD) years	15.8 (4.0)	12.1 (4.0)	10.3 (3.9)	5.6 (4.4)	4.7 (3.9)	<.001
Income, %
Low	20.9	36.7	11.4	53.3	8.2	<.001
Middle	30.2	36.2	57.1	28.5	55.7
High	49.0	27.2	31.5	18.3	36.1
Family physical violence, %
Never or rarely	91.2	96.0	97.2	91.7	96.8	<.001
Sometimes—Frequently	8.8	4.0	2.8	8.3	3.2
Cohabitation, %
Live alone	30.5	24.8	9.4	12.7	5.7	<.001
Live with spouse	63.1	71.1	76.4	50.7	68.4
Live with children	6.5	6.2	42.6	57.0	29.9
Live with family	0.7	3.0	2.8	10.5	42.8
Live with others	18.8	1.0	1.5	2.2	30.3
Smoking, %
Currently	4.8	6.5	12.5	10.5	7.5	<.001
Ex-smoker	51.0	57.2	27.1	43.6	43.3
Never smoked	44.2	36.3	60.4	45.9	49.0
Diabetes, %	13.3	19.2	28.7	13.5	28.6	<.001

Note. IMIAS = International Mobility in Aging Study.

Figure 1.

Frequency of visual impairment in study samples.

Figure 2.

Percentages visiting an eye care professional in the last year.

Associations With Visual Impairment

In the Canadian cities, the 16 subjects with visual impairment had similar age but lower education than those with no visual impairment (mean years of education 11.4 vs. 14, p = .02). Those with visual impairment were more likely to live alone, to live in Saint-Hyacinthe, and to be current smokers (Table 2).

Table 2.

Visual Impairment by Sociodemographic Characteristics and Health Associated Factors in the Canadian Cities and the Non-Canadian Cities.

	n	Canadian cities	p value	n	Non-Canadian cities	p value
Sex			.216			.014
Women	415	1.5		608	18.1
Men	371	2.7		560	12.9
Income			.367			<.001
Low	227	3.1		273	24.2
Middle	259	1.9		555	16.0
High	300	1.3		340	7.9
Physical violence, %			.970			<.001
Never or rarely	735	2.04		1,112	14.7
Sometimes—Frequently	51	2.00		56	32.1
Live alone, %			.040			.003
No	568	1.4		1,061	25.5
Yes	216	3.7		106	14.6
Live with spouse			.045			<.001
No	259	3.5		407	20.9
Yes	527	1.3		761	12.7
Live with children			.309			.971
No	736	1.9		666	15.6
Yes	50	4.0		502	15.5
Live with family			.200			.692
No	771	1.9		944	15.8
Yes	15	6.7		224	14.7
Live with others			.609			.292
No	707	2.1		1,032	16.0
Yes	79	1.3		136	12.5
Research site			.013			<.001
Kingston	390	0.8
Saint-Hyacinthe	396	3.2
Tirana				392	15.1
Manizales				374	22.2
Natal				402	10.0
Smoking			<.001			.02
Current smoker	45	8.9		116	20.7
Ex-smoker	422	2.6		437	11.9
Never smoked	317	0.3		596	17.3
Diabetes			.787			.50
Yes	128	2.0		279	16.8
No	658	2.3		889	15.2

Note. The p value comes from chi-square test.

In the non-Canadian cities, there were 182 (15.6%) subjects with visual impairment. They were older (age 69.1 vs. 70.1, p < .001) and less educated (5.9 vs. 7.1 years of education, p < .001) than those without visual impairment. Being a woman, having a low income, having experienced physical violence from a family member were associated with higher prevalence of visual impairment (Table 2). Living with spouse or children was associated with better vision compared with living alone and other living arrangements. Those living in Manizales and Tirana had higher prevalence of visual impairment than those living in Natal (Table 2); current smokers had higher prevalence of visual impairment. In the fully adjusted model, older adults were more likely to be visually impaired (odds ratio [OR] = 1.12; 95% confidence interval [CI] = [1.09, 1.15]; Table 3). Those who had experienced physical violence from a family member (OR = 1.87; 95% CI = [1.17, 3.00]) were more likely to be visually impaired. Those living with their spouse or with children were less likely to be visually impaired than those living alone or in other living arrangements. Finally those living in Natal were less likely to be visually impaired (OR = 0.39; 95% CI = [0.17, 0.90]) than those residing in Tirana, while there were no differences in visual impairment between Manizales and Tirana once adjustment was carried out for all covariates in the model. Results of the meta-analysis of the fully adjusted association between visual impairment and physical domestic violence showed no heterogeneity of effect across research sites (p = .625), although CIs for each site were large. The estimated site-specific ORs were 2.43 (95% CI = [0.56, 10.47]) in Tirana, 1.56 (95% CI = [0.67, 3.64]) in Manizales, 3.28 (95% CI = [0.88, 13.20]) in Natal, and the pooled association between physical violence and visual impairment in the fully adjusted meta-analysis was 2.02 (95% CI = [1.07, 3.84]).

Table 3.

Adjusted Risk Factors for Visual Impairment From a Multiple Logistic Regression Model: Tirana, Manizales, and Natal.

	Visual impairment OR	95% CI	p value
Age, per year	1.12	[1.09, 1.15]	<.001
Sex (reference: Men)	0.99	[0.68, 1.45]	.977
Education, per year	0.94	[0.88, 1.00]	.076
Income (reference: Low)
Middle	0.78	[0.55, 1.08]	.136
High	0.45	[0.21, 1.00]	.04
Physical violence from family member (reference: Rarely or never)
Sometimes—Frequently	1.87	[1.17, 3.00]	.009
Cohabitation
Live with spouse vs. alone	0.75	[0.62, 0.91]	.003
Live with children vs. alone	0.78	[0.63, 0.98]	.030
Live with family vs. alone	1.34	[0.77, 2.33]	.290
Live with others vs. alone	1.08	[0.58, 2.00]	.805
Research site (reference: Tirana)
Manizales	1.12	[0.87, 1.44]	.349
Natal	0.41	[0.20, 0.87]	.021
Smoking (reference: Never smoked)
Ex-smoker	0.62	[0.40, 0.93]	.022
Current smoker	1.23	[1.08, 1.40]	.002
Diabetes	1.34	[0.71, 2.52]	.370

Note. The p value comes from logistic regression models adjusting for all covariates. CI = confidence interval.

Associations With Visiting an Eye Care Professional in the Last Year

In the Canadian cities, 60% of participants had visited an eye care professional in the last 2 years, with significant differences between cities: Those in Kingston had more frequent visits to eye care professionals (70%) compared with those living in Saint-Hyacinthe (51%; Table 4). Older age and more years of education were significantly associated with more frequent visits to an eye care professional (data not shown). In bivariate analyses, other factors associated with eye care visits in Canada were being a woman and having diabetes. In multivariate analyses, older age, being a woman, having more education, living in Kingston, and having diabetes remained significantly associated with visits to an eye care professional.

Table 4.

Factors Associated With Visiting an Eye Care Professional in the Canadian Cities (Kingston and Saint-Hyacinthe) and the Non-Canadian Cities (Tirana, Manizales, and Natal).

	n	Canadian cities	p value	n	Non-Canadian cities	p value
Sex			.047			.011
Women	420	63.3		605	42.8
Men	374	56.4		565	35.6
Income			.530			.008
Low	228	57.0		277	31.8
Middle	264	60.9		554	40.4
High	302	61.6		339	43.7
Physical violence, %			.237			.050
Never or rarely	744	59.5		1,112	38.7
Sometimes—Frequently	51	68.0		56	51.8
Cohabitation
Live alone, %			.797			.605
No	575	59.8		1,061	39.1
Yes	217	60.8		108	41.7
Live with spouse, %			.458			.132
No	260	61.9		407	42.3
Yes	534	59.1		763	37.7
Live with children, %			.070			.019
No	744	60.9		668	42.2
Yes	50	48.0		502	35.5
Live with family, %			.821			.008
No	780	60.1		947	37.5
Yes	14	57.1		223	47.1
Live with others, %			.781			.007
No	716	59.9		1,034	37.9
Yes	78	61.5		136	50.0
Research site			<.001			<.001
Kingston	396	69.2
Saint-Hyacinthe	398	51.0
Tirana				388	25.8
Manizales				381	37.3
Natal				401	54.4
Smoking			.07			.009
Current smoker	44	45.4		119	27.7
Ex-smoker	428	59.1		446	43.0
Never smoked	320	62.1		604	38.7
Diabetes			<.001			.10
Yes	130	73.8		278	43.5
No	664	57.4		892	38.0

Note. The p value comes from chi-square test.

In the non-Canadian cities, the overall proportion visiting an eye care professional was 39.3%, with significant differences between Natal (54%), Manizales (37%), and Tirana (26%); in bivariate analyses, being a woman, having higher education and income, and having experienced physical violence from a family member were associated with visits to an eye care professional (Table 4). In the fully adjusted model, in addition to sex, education, and income inequalities, which remained significant, history of physical domestic violence was significantly associated with frequency of visits to an eye care professional (PR = 1.35, 95% CI = [1.05,1.72]), and Natal participants used more eye care (Table 5).

Table 5.

Prevalence Ratios for Eye Care Utilization in Last Year From a Poisson Regression Model.

	Canadian cities			Non-Canadian cities
	Prevalence ratios	95% CI	p value	Prevalence ratios	95% CI	p value
Age, per year	1.02	[1.00, 1.04]	.04	0.99	[0.97, 1.02]	.492
Sex (reference: Men)	1.16	[1.03, 1.31]	.02	1.28	[1.10, 1.50]	.002
Education, per year	1.02	[1.00, 1.03]	.008	1.03	[1.01, 1.05]	<.001
Income (reference: Low)
Middle	1.07	[0.92, 1.24]	.376	1.25	[1.02, 1.55]	.035
High	0.99	[0.86, 1.17]	.998	1.32	[1.02, 1.65]	.032
Physical violence (reference: Rarely–never)
Sometimes—Frequently	1.05	[0.92, 1.24]	.628	1.35	[1.05, 1.74]	.017
Research site (reference: Tirana)
Manizales				1.79	[1.42, 2.26]	<.001
Natal				2.65	[2.10, 3.32]	<.001
Research site (reference: Kingston)
Saint-Hyacinthe	0.79	[0.69, 0.90]	.008
Smoking (never smoked)
Ex-smoker	0.98	[0.87, 1.10]	.72	1.13	[0.98, 1.31]	.10
Current smoker	0.77	[0.56, 1.06]	.11	0.83	[0.62, 1.12]	.22
Diabetes	1.39	[1.23, 1.57]	<.001	1.12	[0.96, 1.31]	.15

Note. The p value comes from Poisson’s regression models adjusting for all covariates. CI = confidence interval.

Results obtained from the meta-analysis using the data from the five sites showed that there was no heterogeneity of effect in the fully adjusted association between domestic physical violence and eye care (p = .280) although CIs were large. The estimated site-specific PRs were 1.06 (95% CI = [0.85, 1.33]) in Kingston, 1.07 (95% CI = [0.70, 1.63]) in Saint-Hyacinthe, 1.10 (95% CI = [0.42, 2.88]) in Tirana, 2.43 (95% CI = [0.56, 10.47]) in Manizales, and 1.59 (95% CI = [1.19, 2.12]) in Natal. Pooling the five sites, the measure of association between physical violence and eye care was PR = 1.21 (95% CI = [1.01, 1.44]).

Discussion

Data from the IMIAS have been used to provide information on visual impairment and eye care utilization in older adults in five cities from different contexts of social/societal factors.

Frequency of Visual Impairment and of Eye Care Utilization

Visual impairment according to the WHO definition was infrequent among Canadians and highest in Manizales (Colombia) and Tirana (Albania). There are no peer-reviewed, population-based data on the prevalence of visual impairment using a visual acuity chart in older adults in Albania or Colombia. There is only one Canadian population-based study done in Brantford, Ontario (Robinson et al., 2013). Our prevalence in Natal, Brazil (10%) was somewhat higher than what was reported in the São Paulo Eye Study which found a prevalence of presenting visual impairment of 6.2% in adults aged 50 years and older (Salomao et al., 2008), but this prevalence difference could be explained by the older age and the lower socioeconomic status of the IMIAS Natal sample.

Visual impairment was more frequent where eye care utilization was lower both in the cities outside of Canada and in the Canadian cities. Comparisons of eye care utilization between Natal (Brazil) and Manizales (Colombia) and Tirana (Albania) are illustrative. In 1999, Brazil made a national effort to increase access to modern cataract surgery by increasing personnel, equipment, and supplies in the public hospitals (Salomao et al., 2009). This resulted in a tripling of the rate of cataract surgery from 600 surgeries per million of population in the years immediately prior to 1998 to 1,815 surgeries per million of population in 2002 (Salomao et al., 2009). Data from the Brazilian government obtained from 2012 show that 432,000 cataract surgeries were performed free of charge through “The Program of Elective Surgeries” with a rate of 2,171 surgeries per million (Saúde, 2013). Furthermore, we found that 54% of older adults in Natal reported seeing an eye care provider in the last year. This is somewhat higher than what was reported in 2012 from a cluster-based sample of urban and rural adults aged 60 years and older in Brazil using data from the World Health Survey in which 30% to 40% reported seeing an eye care provider in the last year (Vela et al., 2012). By contrast, the health care systems in Colombia and Albania struggle with underfunding and high out-of-pocket expenditures (Ylli, 2010). In our data, less than 40% of older adults in Manizales and Tirana had seen an eye care provider in the last year. The lower use of eye care in Manizales and Tirana is consistent with the lower coverage of health care in those two cities and their higher visual impairment.

Significant differences in utilization of eye care were found between the two Canadian study sites. The lower use of eye care in Saint-Hyacinthe is consistent with our previous findings of less use of medical care in that city as shown by their lower control of hypertension (Doulougou et al., 2016) and their lower use of anti-depressive medication (Nana et al., 2015).

Sociodemographic Factors and Sex Were Associated With Visual Impairment and More Frequent Eye Care

Previous studies have found relationships between low education and visual impairment (Cockburn et al., 2012; Freeman et al., 2013; Jin & Wong, 2008). The association between female gender and visual impairment has already been reported in previous studies in the developing world (Zetterberg, 2016). This association was not significant in the Canadian populations in our study, and while women from Manizales, Tirana, and Natal did have a higher prevalence of visual impairment than men, this association disappeared after adjustment for education, income, exposure to family physical violence, and living arrangements, indicating that the excess prevalence of visual impairment in women living in those cities is probably related to gender differences of social origin. Outside of Canada, visual impairment was associated with living alone or living arrangements other than spouse or children. This difference reflects the central role of immediate family in the lives of Mediterranean and Latin American older adults.

Greater education was associated with eye care utilization in the last year. This is in agreement with previous research (Orr, Barron, Schein, Rubin, & West, 1999; Schaumberg, Christen, Glynn, & Buring, 2000) but it is noteworthy that this finding in Canada was independent of income suggesting that education acts as a factor predisposing to eye care. While income is not a barrier to eye care in our Canadian populations, in Manizales, Tirana, and Natal, poor participants used less care, demonstrating the social inequalities still present in the health care systems in those cities (Vela et al., 2012). Women used more eye care than men in spite of having the same adjusted prevalence of visual impairment. The literature is not consistent with this finding as women have less coverage of eye care and cataract surgery in many low-income countries (Mganga, Lewallen, & Courtright, 2011).

Family Physical Violence as a Risk Factor for Visual Impairment and More Frequent Eye Care in Old Age

A recent study from Brazil reports on the high frequency of facial injuries in older adults who report family physical violence (de Sousa et al., 2016). Physical violence can directly lead to eye injuries such as hyphema, corneal abrasions, retinal detachments, globe ruptures, chemical burns, and orbital fractures (Beck, Freitag, & Singer, 1996; Brian, du Toit, Ramke, & Szetu, 2011; Hartzell, Botek, & Goldberg, 1996). Unfortunately, we did not specifically ask questions about eye injuries, and it was beyond the scope of this study to perform a comprehensive ophthalmological exam. Typically, eye injuries only affect one eye which would not have an immediate effect on binocular visual acuity, although monocular visual impairment does increase the risk of subsequent binocular visual impairment (Varma et al., 2010). Sometimes, a domestic violence-related eye injury can affect both eyes at the same time, such as in the case of a chemical burn (Beare, 1990). As physical violence can result in eye injuries, it may also be associated with eye care utilization.

Associations Between Diabetes and Eye Care

Having diabetes is associated with eye care utilization in the Canadian cities and in Manizales which may indicate better diabetes care at these cities. Diabetes is becoming a leading cause of visual impairment and blindness in adults (Glover et al., 2012; Leasher et al., 2016), and given the current global pandemic of diabetes it is extremely important that patients with diabetes see an eye care provider on a regular basis, so that they can obtain timely treatment for diabetic retinopathy.

Limitations of this study include its cross-sectional design, the lack of data on the pathological mechanism of vision loss or reports of previous eye injuries, and the examination of other measures of visual function such as visual field. Also, our response rates were low in the Canadian sites. However, a comparison of our samples with the Canadian census sample for older adults in Saint-Hyacinthe indicates that participants are representative of the underlying population in education, marital status, and the proportion of the sample under the official threshold for low income. The sample of Kingston is overeducated but still representative of marital status and the proportion of people under the official Canadian threshold for low income. Finally, there remains a significant prospect of residual confounding due to lack of information on family history of eye disease, ethnicity, and pathologies associated with visual impairment.

Implications

Based on data from other eye studies done in Brazil and elsewhere in the developing world, it is highly probable that cataract and uncorrected refractive errors, both of which are highly treatable, are the two main causes of the visual impairment that we identified (Resnikoff, Pascolini, Mariotti, & Pokharel, 2008). The inability to see can make it difficult for an older adult to do essential tasks such as take medications properly, fix a meal, get exercise, pay the bills, find employment, and live independently (Harrabi, Aubin, Zunzunegui, Haddad, & Freeman, 2014). Their mobility can be impaired, and they may be unable to drive themselves. Some research shows that visual impairment is related to cognitive impairment and a greater risk of death (Freeman, Egleston, West, Bandeen-Roche, & Rubin, 2005). Yet, research shows that the leading causes of visual impairment in the developing world are treatable (Pascolini & Mariotti, 2012).

To conclude, high frequency of visual impairment and poor eye care utilization were found in Manizales (Colombia) and Tirana (Albania). Exposure to family physical violence emerged as a potential risk factor for visual impairment and increased eye care use in old age, although the role of physical violence in age-related vision loss needs to be further examined. Finally, public policies to decrease social inequalities and domestic violence could increase the opportunities for active aging and good vision in old age.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Canadian Institutes of Health Research funded the study (AAM 108751).

References

Beare

J. D.

(1990). Eye injuries from assault with chemicals. British Journal of Ophthalmology, 74, 514-518.

Beck

S. R.

Freitag

S. L.

Singer

(1996). Ocular injuries in battered women. Ophthalmology, 103, 148-151.

Brian

du Toit

Ramke

Szetu

(2011). Population-based study of self-reported ocular trauma in Fiji. Clinical & Experimental Ophthalmology, 39, 441-448. doi:10.1111/j.1442-9071.2010.02476.x

Cockburn

Steven

Lecuona

Joubert

Rogers

Cook

Polack

(2012). Prevalence, causes and socio-economic determinants of vision loss in Cape Town, South Africa. PLoS ONE, 7(2), Article e30718. doi:10.1371/journal.pone.0030718

de Sousa

R. I.

de Macedo Bernardino

Castro

R. D.

Cavalcanti

A. L.

Bento

P. M.

d’Avila

. (2016). Facial trauma as physical violence markers against elderly Brazilians: A comparative analysis between genders. Archives of Gerontology and Geriatrics, 67, 55-60. doi:10.1016/j.archger.2016.06.015

De Yebenes

M. J.

Otero

Zunzunegui

M. V.

Rodriguez-Laso

Sanchez-Sanchez

Del Ser

. (2003). Validation of a short cognitive tool for the screening of dementia in elderly people with low educational level. International Journal of Geriatric Psychiatry, 18, 925-936. doi:10.1002/gps.947

Doulougou

Gomez

Alvarado

Guerra

R. O.

Ylli

Guralnik

Zunzunegui

M. V.

(2016). Factors associated with hypertension prevalence, awareness, treatment and control among participants in the International Mobility in Aging Study (IMIAS). Journal of Human Hypertension, 30, 112-119. doi:10.1038/jhh.2015.30

Fenwick

E. K.

Ong

P. G.

Man

R. E.

Cheng

C. Y.

Sabanayagam

Wong

T. Y.

Lamoureux

E. L.

(2016). Association of vision impairment and major eye diseases with mobility and independence in a Chinese population. JAMA Ophthalmology, 134, 1087-1093. doi:10.1001/jamaophthalmol.2016.2394

Freeman

E. E.

Egleston

B. L.

West

S. K.

Bandeen-Roche

Rubin

(2005). Visual acuity change and mortality in older adults. Investigative Ophthalmology & Visual Science, 46, 4040-4045. doi:10.1167/iovs.05-0687

10.

Freeman

E. E.

Roy-Gagnon

M. H.

Samson

Haddad

Aubin

M. J.

Vela

Zunzunegui

M. V.

(2013). The global burden of visual difficulty in low, middle, and high income countries. PLoS ONE, 8(5), Article e63315. doi:10.1371/journal.pone.0063315

11.

Glover

S. J.

Burgess

P. I.

Cohen

D. B.

Harding

S. P.

Hofland

H. W.

Zijlstra

E. E.

Allain

T. J.

(2012). Prevalence of diabetic retinopathy, cataract and visual impairment in patients with diabetes in sub-Saharan Africa. British Journal of Ophthalmology, 96, 156-161. doi:10.1136/bjo.2010.196071

12.

Guedes

D. T.

Alvarado

B. E.

Phillips

S. P.

Curcio

C. L.

Zunzunegui

M. V.

Guerra

R. O.

(2015). Socioeconomic status, social relations and domestic violence (DV) against elderly people in Canada, Albania, Colombia and Brazil. Archives of Gerontology and Geriatrics, 60, 492-500. doi:10.1016/j.archger.2015.01.010

13.

Guedes

D. T.

Vafaei

Alvarado

B. E.

Curcio

C. L.

Guralnik

J. M.

Zunzunegui

M. V.

Guerra

R. O.

(2016). Experiences of violence across life course and its effects on mobility among participants in the International Mobility in Aging Study. BMJ Open, 6(10), Article e012339. doi:10.1136/bmjopen-2016-012339

14.

Harrabi

Aubin

M. J.

Zunzunegui

M. V.

Haddad

Freeman

E. E.

(2014). Visual difficulty and employment status in the world. PLoS ONE, 9(2), Article e88306. doi:10.1371/journal.pone.0088306

15.

Hartzell

K. N.

Botek

A. A.

Goldberg

S. H.

(1996). Orbital fractures in women due to sexual assault and domestic violence. Ophthalmology, 103, 953-957.

16.

Jin

Y. P.

Wong

D. T.

(2008). Self-reported visual impairment in elderly Canadians and its impact on healthy living. Canadian Journal of Ophthalmology, 43, 407-413. doi:10.3129/i08-077

17.

Krishnaiah

Vilas

Shamanna

B. R.

Rao

G. N.

Thomas

Balasubramanian

(2005). Smoking and its association with cataract: Results of the Andhra Pradesh eye disease study from India. Investigative Ophthalmology & Visual Science, 46, 58-65. doi:46/1/58 [pii] 10.1167/iovs.04-0089

18.

Leasher

J. L.

Bourne

R. R.

Flaxman

S. R.

Jonas

J. B.

Keeffe

Naidoo

. . .Taylor

H. R.

(2016). Global estimates on the number of people blind or visually impaired by diabetic retinopathy: A meta-analysis from 1990 to 2010. Diabetes Care, 39, 1643-1649. doi:10.2337/dc15-2171

19.

Maberley

D. A.

Hollands

Chuo

Tam

Konkal

Roesch

. . .Bassett

(2006). The prevalence of low vision and blindness in Canada. Eye, 20, 341-346. doi:10.1038/sj.eye.6701879

20.

Mganga

Lewallen

Courtright

. (2011). Overcoming gender inequity in prevention of blindness and visual impairment in Africa. Middle East African Journal of Ophthalmology, 18, 98-101. doi:10.4103/0974-9233.80695

21.

Nana

G. N.

Doulougou

Gomez

Ylli

Guralnik

Zunzunegui

M. V.

(2015). Social differences associated with the use of psychotropic drugs among men and women aged 65 to 74 years living in the community: The International Mobility in Aging Study (IMIAS). BMC Geriatrics, 15, Article 85. doi:10.1186/s12877-015-0083-3

22.

Orr

Barron

Schein

O. D.

Rubin

G. S.

West

S. K.

(1999). Eye care utilization by older Americans: The SEE project. Salisbury Eye Evaluation. Ophthalmology, 106, 904-909.

23.

Pascolini

Mariotti

S. P.

(2012). Global estimates of visual impairment: 2010. British Journal of Ophthalmology, 96, 614-618. doi:10.1136/bjophthalmol-2011-300539

24.

Rabe-Hesketh

Skrondal

(2008). Multilevel and longitudinal modeling using Stata. College Station, Texas: Stata Press.

25.

Resnikoff

Pascolini

Etya’ale

Kocur

Pararajasegaram

Pokharel

G. P.

Mariotti

S. P.

(2004). Global data on visual impairment in the year 2002. Bulletin of the World Health Organization, 82, 844-851. doi:S0042-96862004001100009

26.

Resnikoff

Pascolini

Mariotti

S. P.

Pokharel

G. P.

(2008). Global magnitude of visual impairment caused by uncorrected refractive errors in 2004. Bulletin of the World Health Organization, 86, 63-70. doi:S0042-96862008000100017

27.

Robinson

Feng

Woods

C. A.

Fonn

Gold

Gordon

(2013). Prevalence of visual impairment and uncorrected refractive error—Report from a Canadian urban population-based study. Ophthalmic Epidemiology, 20, 123-130. doi:10.3109/09286586.2013.789915

28.

Salomao

S. R.

Cinoto

R. W.

Berezovsky

Araujo-Filho

Mitsuhiro

M. R.

Mendieta

. . .Ellwein

L. B.

(2008). Prevalence and causes of vision impairment and blindness in older adults in Brazil: The Sao Paulo Eye Study. Ophthalmic Epidemiology, 15, 167-175. doi:10.1080/09286580701843812

29.

Salomao

S. R.

Soares

F. S.

Berezovsky

Araujo-Filho

Mitsuhiro

M. R.

Watanabe

S. E.

. . .Ellwein

L. B.

(2009). Prevalence and outcomes of cataract surgery in Brazil: The Sao Paulo eye study. American Journal of Ophthalmology, 148, 199-206.e2. doi:10.1016/j.ajo.2009.02.019

30.

Saúde

M. d.

(2013). É tempo a diminuir a espera. É tempo de saúde [Reducing waiting time, increasing healthy time]. Brazil. Retrieved from http://bvsms.saude.gov.br/bvs/publicacoes/ministerio_municipios_juntos_acesso_integral.pdf

31.

Schaumberg

D. A.

Christen

W. G.

Glynn

R. J.

Buring

J. E.

(2000). Demographic predictors of eye care utilization among women. Medical Care, 38, 638-646.

32.

Schellini

S. A.

Durkin

S. R.

Hoyama

Hirai

Cordeiro

Casson

R. J.

Padovani

C. R.

(2009). Prevalence and causes of visual impairment in a Brazilian population: The Botucatu Eye Study. BMC Ophthalmology, 9, Article 8. doi:10.1186/1471-2415-9-8

33.

Sherin

K. M.

Sinacore

J. M.

X. Q.

Zitter

R. E.

Shakil

(1998). HITS: A short domestic violence screening tool for use in a family practice setting. Family Medicine, 30, 508-512.

34.

Sousa

A. C.

Guerra

R. O.

Thanh Tu

Phillips

S. P.

Guralnik

J. M.

Zunzunegui

M. V.

(2014). Lifecourse adversity and physical performance across countries among men and women aged 65-74. PLoS ONE, 9(8), Article e102299. doi:10.1371/journal.pone.0102299

35.

Stevens

G. A.

White

R. A.

Flaxman

S. R.

Price

Jonas

J. B.

Keeffe

. . .Bourne

R. R.

(2013). Global prevalence of vision impairment and blindness: Magnitude and temporal trends, 1990-2010. Ophthalmology, 120, 2377-2384. doi:10.1016/j.ophtha.2013.05.025

36.

Taylor

H. R.

(1978). Applying new design principles to the construction of an illiterate E chart. American Journal of Optometry and Physiological Optics, 55, 348-351.

37.

United Nations Development Programme. (2015). Human development reports. Retrieved from http://hdr.undp.org/en/composite/GII

38.

Varma

Chung

Foong

A. W.

Torres

Choudhury

Azen

S. P.

, . . . Los Angeles Latino Eye Study Group. (2010). Four-year incidence and progression of visual impairment in Latinos: The Los Angeles Latino Eye Study. American Journal of Ophthalmology, 149, 713-727. doi:10.1016/j.ajo.2009.12.011

39.

Vela

Samson

Zunzunegui

M. V.

Haddad

Aubin

M. J.

Freeman

E. E.

(2012). Eye care utilization by older adults in low, middle, and high income countries. BMC Ophthalmology, 12, Article 5. doi:10.1186/1471-2415-12-5

40.

Williams

R. L.

(2000). A note on robust variance estimation for cluster-correlated data. Biometrics, 56, 645-646.

41.

Ylli

(2010). Health and social conditions of older people in Albania: Baseline data from a national survey. Public Health Reviews, 32, 549-560.

42.

Zetterberg

(2016). Age-related eye disease and gender. Maturitas, 83, 19-26. doi:10.1016/j.maturitas.2015.10.005

43.

Zunzunegui

M. V.

Alvarado

B. E.

Guerra

Gomez

J. F.

Ylli

Guralnik

J. M.

(2015). The mobility gap between older men and women: The embodiment of gender. Archives of Gerontology and Geriatrics, 61, 140-148. doi:10.1016/j.archger.2015.06.005

44.

Zunzunegui

M. V.

Beland

Otero

(2001). Support from children, living arrangements, self-rated health and depressive symptoms of older people in Spain. International Journal of Epidemiology, 30, 1090-1099.