Study sample and procedures

The data were obtained from the PolSenior2 study conducted in 2018 and 2019 as a cross‑sectional survey of a representative sample of 5987 noninstitutionalized participants aged 60 years and over. The participants were recruited from all administrative regions in Poland using a 3‑stage stratified, proportional draw. The participants were split into 5‑year age intervals: 60 to 64, 65 to 69, 70 to 74, 75 to 79, 80 to 84, 85 to 89, and equal to or above 90 years. The cohorts were similar in terms of size and sex distribution. The study protocol consisted of 3 questionnaires, specific geriatric scales, tests, as well as anthropometric and blood pressure measurements. Medical and socioeconomic survey data were collected through face‑to‑face interviews conducted by trained nurses during 3 home visits with the participants. The remaining data were collected through a self‑completion questionnaire that the respondents filled out individually.¹³ Initial ophthalmology screenings were carried out during the first visit. In the present analysis, data from 5831 individuals were used. A small percentage of respondents (2.2%) were excluded as they did not undergo near visual acuity assessment, but the lack of information was random. All the individuals included in the study signed a written informed consent form. The study protocol was approved by the Bioethics Committee of the Medical University of Gdansk (NKBBN/257/2017).

Visual impairment assessment

Binocular near vision with habitual correction was evaluated using the Snellen chart for near vision, according to the vision correction customarily worn by the patient. If a participant habitually wore eyeglasses in their daily life, we assessed their vision with their eyeglasses on.

The chart consisted of 8 sections of text with gradually increasing font size, marked from 1 (smallest font) to 8 (largest font). Functional near vision was measured at each participant’s preferred distance. For illiterate individuals, an “E” chart was utilized. The patients who were unable to read any lines were instructed to count fingers at a distance of 30 to 40 cm. Among the patients whose visual acuity was below finger count, light perception was evaluated using the question “Can you distinguish between brightness and darkness?” Blindness was diagnosed if the answer to that question was “No.” Additionally, the patients were asked if they were able to watch television.

Based on the above data, the respondents were categorized into the following groups: 1) normal vision, including individuals who could read sections 1 to 4 from the standard distance (30–40 cm) and were able to watch television; 2) visual impairment, including individuals who could not read sections 1 to 4 from the standard distance (30–40 cm) or sections 5 to 8 from any distance, were unable to read any text / count fingers, had no light perception, or could not watch television.¹⁴

Socioeconomic factors

To assess the prevalence of visual impairment, we evaluated the following socioeconomic factors among men and women aged 60 years and above: education level, financial situation, marital status, living arrangements, and occupational group.

Data on the patient socioeconomic factors were obtained from the abovementioned socioeconomic survey. To determine their education level, the respondents were presented with the question “What is your current education level?” The possible answers were 1) no education, 2) incomplete primary school, 3) primary school, 4) secondary school, 5) vocational high school, 6) high school, 7) 2‑year college, 8) bachelor’s degree, and 9) master’s degree. To simplify the analysis, education levels 1 to 3 were classified as “primary,” 4 to 6 were classified as “secondary,” and 7 to 9 were classified as “higher.”

The participants’ financial situation was evaluated with the question “Which of the following sentences best describes the financial situation of your household?” The response options were 1) “I / we live comfortably without having to save for special purchases,” 2) “I / we live within our means and have enough money for all our needs,” 3) “I / we need to put money aside to save for special purchases,” 4) “I /we have enough money for basic needs, such as food and clothing,” 5) “I / we only have enough money for food,” and 6) “I / we do not have enough money to meet our basic needs.” For result consistency, responses were grouped as follows: “Can easily afford everything” (1), “Can afford everything but only when saving” (2), and “Financial difficulties” (3–6).

Marital status was determined by the question “How would you describe your current marital status?” There were 4 answers: 1) “Never been married,” 2) “I am married,” 3) “I am a widower / widow,” and 4) “I am divorced / separated.” Answers 1, 3, and 4 were classified as “single” while the answer 2 was classified as “married.”

To determine the participants’ living situation, 2 questions were used: “Do any other people live in the apartment / house with you?” A negative response was coded as “living alone.” If the respondent answered “yes,” they were asked a second question: “Who do you live with?” The possible answers comprised 1) husband / wife (including former) or partner(s), which was coded as “with spouse only,” and 2) children, 3) grandchildren, 4) great‑grandchildren, 5) parents‑in‑law, 6) parents, 7) other family members, 8) people outside the family, all of which were classified as “others.”

The respondents were asked to specify the occupational group they belonged to for the longest period of their life (white‑collar, manual / farming, other, or none) to determine their occupation history.

Medical factors

We also assessed the prevalence of comorbid conditions, including arterial hypertension, diabetes, and obesity in the participants with and without visual impairment.

Diabetes mellitus was considered present when a patient declared that it was previously diagnosed, if the fasting glucose level was greater than or equal to 126 mg/dl, or if the use of hypoglycemic drugs was reported.

Hypertension was diagnosed if the average blood pressure values from 2 measurements during each visit were equal to or greater than 140 mm Hg (systolic blood pressure) and / or 90 mm Hg (diastolic blood pressure) or if the patient was taking hypotensive drugs over the preceding 2 weeks because of an earlier diagnosis of hypertension.

Obesity was defined as body mass index (BMI) greater than or equal to 30 kg/m², calculated using an approved Tanita BC‑545 N portable electronic scale (Medkonsulting Sp. z o. o., Poznań, Poland), with an accuracy of 0.1 kg.

Statistical analysis

Statistical analyses were performed using SAS 9.4 TS Level 1M5 (SAS Institute, Cary, North Carolina, United States) and R version 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria). The results were presented as percentages with 95% CIs. Sampling weights were included in statistical calculations to account for the complex survey design using the R “survey” package. The poststratification procedure was used to match the age–sex sample distribution to the population of Poland. Univariable and multivariable logistic regression models were created to identify socioeconomic risk factors for visual impairment. All variables for which prevalence was presented in Table 1 were included in the multivariable model. For all statistical analyses, the level of significance was set at 0.05.

Patient characteristics

The study sample was comprised of 5855 individuals at a mean (SD) age of 74.8 (9.4) years (range, 60–106 years). Women comprised 50.8% of the study population. Overall, nearly 35.3% of the total sample lived in rural areas. Approximately 28.7% of the participants reported a primary level of education, which included elementary education, incomplete elementary education, or no formal education. A secondary level of education (secondary school, vocational high school, or high school) was declared by 53.9% of the participants, while 17.4% reported having higher education, which included 2‑year college, bachelor’s degree, and master’s degree. Additionally, 39.2% were single, including widows / widowers and individuals who were divorced or separated from their spouses. The distributions of education level, age, sex, marital status, employment type, self‑reported poverty, and living situation in the study sample are shown in Table 1. The group of participants who were still active workers was very small, comprising only 7.1% of the total cohort (3.7% of all women and 10.5% of all men). The prevalence of visual impairment did not differ between those not working (43.1%) and active workers (32.9%) (Table 2). A total of 55.8% (95% CI, 53.7%–58%) of the participants wore eyeglasses. Usage of eyeglasses was less frequent among the individuals with dementia, as compared with those without (46.5% [95% CI, 41.7%–51.4%] vs 57.5% [95% CI, 55.2%–59.8%]). Additionally, eyeglasses were worn less frequently by individuals with a primary than those with a secondary level of education (50.6% [95% CI, 46.7%–54.5%] vs 57.3% [95% CI, 54.7%–59.9%]). No significant difference in eyeglasses usage was found according to the financial situation of the household.

Socioeconomic risk factors for visual impairment

The prevalence of visual impairment in the older Polish adult population was estimated at 41.9% (95% CI, 38.3%–45.4%) among women and 41.7% (95% CI, 37.4%–45.9%) among men. To ensure representativeness, the estimates were weighted and the complex sampling design was taken into account.

Furthermore, the prevalence of visual impairment showed an increase with age and was found to be higher among the individuals with primary education and those engaged in manual labor or farming in the past, as shown in Table 2.

Univariable and multivariable logistic regression analyses

In univariable logistic regression analysis, higher prevalence of visual impairment was found among the respondents aged 75 years and older, those with secondary or primary education levels, individuals who self‑reported poverty, those who were single, residents of rural areas, and those engaged in blue‑collar work, as detailed in Table 3. In multivariable regression analysis, the odds of visual impairment were found to be 1.72‑fold higher (95% CI, 1.52–1.94; P <⁠0.001) among the individuals aged 75 years and over than in those aged 60 to 74 years. The respondents who only completed their primary education had 1.66‑fold greater odds of visual impairment than those with higher education (95% CI, 1.33–2.06; P <⁠0.001). Moreover, the odds of visual impairment were 1.25‑fold higher (95% CI, 1.07–1.47) for individuals engaged in manual / farming work in the past, 1.42‑fold higher (95% CI, 1.11–1.81) for those not engaged in any work in the past, and 1.26‑fold higher (95% CI, 1.02–1.56) for those previously engaged in other types of work, as compared with white‑collar workers. All of these associations were significant (Table 4). Additionally, obesity was associated with a 0.85‑fold lower odds of visual impairments (95% CI, 0.75–0.96; P = 0.007).

Conclusions

We showed that visual impairment was the most prevalent among older age groups (≥75 years), individuals with lower education levels or no education, the unemployed, and those who engaged in manual labor, farming, or other types of non–white‑collar work during their occupational career. Additionally, obesity was inversely associated with the odds of visual impairment. These findings can be used to develop interventions and public health initiatives aimed at addressing visual health disparities, particularly among vulnerable populations, such as the elderly and those with limited educational and employment experience.