Article Text

Download PDFPDF

Economic inequality in presenting near vision acuity in a middle-aged population: a Blinder–Oaxaca decomposition
  1. Mohammad Hassan Emamian1,
  2. Hojjat Zeraati2,
  3. Reza Majdzadeh2,3,
  4. Mohammad Shariati4,
  5. Hassan Hashemi5,6,
  6. Ebrahim Jafarzadehpur7,
  7. Akbar Fotouhi2
  1. 1School of Public Health, Shahroud University of Medical Sciences, Shahroud, Iran
  2. 2Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran
  3. 3Knowledge Utilization Research Centre (KURC), Tehran University of Medical Sciences, Tehran, Iran
  4. 4Department of Community Medicine, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
  5. 5Noor Ophthalmology Research Centre, Noor Eye Hospital, Tehran, Iran
  6. 6Farabi Eye Hospital, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
  7. 7Department of Optometry, Tehran University of Medical Sciences, Tehran, Iran
  1. Correspondence to Professor A Fotouhi, Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran 14155-6446, Iran; afotouhi{at}


Purpose To investigate economic inequality and its determinants in near vision, in a middle-aged population.

Methods In this population-based study, the presenting near vision in the better eye was investigated as the main outcome. If this vision revealed a value ≥1.6 M, the participant was considered to have presenting near vision impairment (PNVI). The effect of variables studied on PNVI was investigated using logistic regression. Economic inequality in PNVI was investigated using the Oaxaca–Blinder decomposition method.

Results PNVI in the better eye was observed in 18.2% (95% CI 16.8 to 19.6) of participants. The prevalence of PNVI in the high and low economic groups was 11.7% (95% CI 10.3 to 13.0) and 28.5% (95% CI 26.0 to 31.0), respectively. Age and education were observed as the main factors in the explained portion of this gap and were in favour of the high economic group. Gender and eye care utilisation were factors affecting the unexplained portion of this gap and were in favour of the low economic group.

Conclusions Economic inequality plays a significant role in PNVI, while age and education are among the main factors affecting this gap. The effect caused by any change in these two factors was found to have a greater effect on women with low economic status.

  • Epidemiology
  • Public health

Statistics from


Statistics reveal an alarming population of more than 1 billion people affected by functional presbyopia globally, of whom 517 million do not have appropriate correction for near vision. Also, 410 million of these suffer from considerable disability1 although the refractive errors are easily correctable. Correction of refractive errors by prescribing glasses is one of the most simple and cost effective interventions in eye care.2 Today, 67% of people with presbyopia and 94% of the population possessing significant near vision disability live in the less or least developed countries.1 Previous studies have shown that the degree to which people use spectacles and the prevalence of refractive errors in urban and rural areas is not the same.2 ,3 Also, blindness is found to be more prevalent, along with a corresponding low percentage of spectacle users and those undergoing cataract surgery, in poor or low economic communities.4–6

Gross inequalities in health, both between and within countries, are a challenge for public health.7 Across all levels of income and in all countries, the lower socioeconomic status is related to a lower level of general health.8 To tackle health inequalities, policy makers in the health and other sectors of governments must be aware and understand the role of social determinants on health. Researchers must expand the scope of evidence across the population and by regular review of the research gap they must find the research needs.8

Although it can be presumed that visual disorders are more prevalent among the low economic groups of society, to the best of our knowledge the degree of this inequality in near vision impairment has not yet been reported. Consequently, the determinants of this inequality, which play an important role in appropriate policy making, have also not been investigated. Given that 65% of visually impaired people and 82% of those who are blind are ≥50 years of age,9 it appears that further studies on eye problems, particularly near vision disorders, in these age groups are warranted.

This population based study investigated the economic inequality and its determinants in near vision in an urban population of Iran.


The first phase of the Shahroud Eye Cohort Study was conducted on the age group 40–64 years in the city of Shahroud, northern Iran, in 2009. The protocol of study has been reported previously.10 In brief, participants were selected randomly and received thorough ophthalmology and optometry examinations. Far refractive errors were measured using the Topcon KR 8800 autorefractometer and retinoscopy. Corrected and uncorrected far and near visual acuity were then measured at 4 m and 40 cm distances, respectively, employing standard charts. In this study, the presenting near vision was investigated as the main outcome. To evaluate this vision, if the participants used near spectacles, his/her corrected near vision with the current spectacles was considered, and if he/she did not use near vision spectacles, his/her uncorrected near vision was considered. To involve the status of both eyes in the analyses, presenting vision in the better eye was used as the outcome variable. If the presenting near vision in the better eye of an individual was ≥1.6 M, the individual was considered to have near vision impairment. Participants with this vision impairment were studied under the different categories of age, gender, economic status, education, insurance and eye care utilisation, and 95% CIs were calculated for the estimates. The effect of the independent variables on the outcome variable was investigated using simple and multiple logistic regression models.

Given that there is no acceptable standard to classify society into different income groups in developing countries, the principal component analysis was done on the participants’ home assets, and then a new variable was created. Finally, this new variable was divided into three tertiles. The first and third tertiles were considered the high and low economic groups, respectively.

Difference in the prevalence of near vision impairment in both the high and low economic groups was decomposed to its components using the Oaxaca–Blinder decomposition method. In brief, this method breaks down the gap between the outcome variable into two components, in both groups. One component results from the apparent differences in the determinants (termed the explained or endowment component) such as differences in educational level between both groups; the other component results from the differences between the two groups with respect to the degree of the effect of these determinants (termed the unexplained or coefficient component).11

To perform the decomposition, the Oaxaca command in STATA software V.11 was used.12 The effect of cluster sampling was also included in the calculation of CIs. In all of the tests, the significance level of 0.05 (95% CI) was used.


In this study, 5190 of the invitees (82.2%) participated in the study. Measurements of presenting near vision for 5128 participants in the right eye and for 5105 people in the left eye were recorded. The prevalence of presenting near vision impairment (PNVI) for the right, left and better eye were 23.6% (95% CI 22.1 to 25.2), 22.9% (95% CI 21.4 to 24.4) and 18.2% (95% CI 16.8 to 19.6), respectively. Table 1 shows this prevalence in the different age, gender, economic, education, insurance and eye care utilisation groups. The difference in the prevalence of PNVI was significant in the different levels of the variables mentioned, except for insurance.

Table 1

Presenting near visual impairment and its determinants in Shahroud, Iran, 2009

The multiple logistic regression model revealed that all of the variables influenced PNVI, although to varying degrees. Economic status was one of the most important variables in this model, indicating that the prevalence of PNVI in those participants from the low economic group was significantly higher than those from the high economic group. In fact, there was a significant gap between these two groups.

As observed in table 2, the difference in the prevalence of PNVI between the two groups (high and low economic status) was 16.8. This gap was observed to favour those from the high economic group. The results of the decomposition showed that 74% (12.4/16.8) of this gap was related to the explained portion while the rest (26%) was related to the unexplained portion. Age and education were the most important factors creating this gap between the two groups.

Table 2

Oaxaca decomposition of near visual impairment gap between economic groups in Shahroud, 2009


This study showed that there was a significant gap in the prevalence of PNVI between the economic groups studied. To a large extent this gap (74%) was found to be related to the explained portion, which implies that this part of the gap could be eliminated if the amount of the investigated variables in the low economic group became similar to that of the high economic group. For example, the average number of years of education in the groups with high and low economic status were 9.6 and 4.3 years (p<0.001), respectively. If both groups were subjected to the same conditions, then 70% (8.7/12.4) of the explained portion would disappear. Other studies from China13 and India14 have considered education as a factor that exerts no effect on presbyopia, while in Iran15 and Tanzania16 presbyopia was seen to increase with an increase in educational level. Table 2 shows that an increase in education results in a decrease in the prevalence of PNVI. Therefore, although the educated population is more likely to suffer from presbyopia, they will probably have corrected it due to the need to use near vision spectacles, and consequently will exhibit a lower degree of PNVI. However, while the illiterate population or the less educated population have a lower occurrence of presbyopia, they show a higher degree of PNVI compared with the literate population, due to neglect or not correcting their PNVI. While reviewing different articles, it should be noted that the results will be different based on sample size, age group, type of variables in the regression models, etc. Concerning age, which plays a significant role in causing this gap, 18.5% (2.3/12.4) of the explained portion of the gap will decrease if the two groups experience the same conditions. Mean age for the high economic status group was 49.8 years compared with 52.2 years for the low economic group (p<0.001). As aging is associated with an increase in presbyopia and a decrease in near vision acuity,13 ,15 ,17 assimilating the age of the two groups could imply that it is better to concentrate on giving economic assistance to the upper age groups of society. This is because these groups not only have lower economic status, they also possess different types of diseases, especially chronic and non-communicable diseases, which are more prevalent at this age.

The present insurance of participants exerted no significant effect on the gap between the two groups. In fact, 94.1% of the population studied was covered by basic insurance; however, a few optometry services such as prescriptions for spectacles are not covered by such types of insurance. This could explain the minimal effect of basic insurance on the gap studied. Similar to insurance, eye care utilisation was also observed to show very little effect (8%) on the gap between the two groups in the explained portion. However, this variable had a significant effect in the unexplained portion. The negative value of this variable in the explained portion (favouring the low economic group) implies that the group with a lower economic status is more influenced by changes in eye care utilisation than the group with a high economic status. We have already shown that 5.7% of the participants of this study had an unmet need in visual acuity,18 and a significant percentage will certainly exhibit the same condition for near vision as well. Thus it can be concluded that the integration of optometry (ophthalmology) examinations in the ‘family physician’ programme currently running across Iran19 is an essential measure to achieve equality in visual health.

The prevalence of PNVI was not seen to differ significantly between the two genders (table 1). Most of the studies on presbyopia have reported a higher prevalence in women.14–17 We found that PNVI was significantly lower in women (table 1). Therefore, it can be concluded that although the prevalence of presbyopia is higher in women, with an appropriate correction, they show better PNVI than men. Gender also becomes significant in the unexplained portion, as seen in table 2, with a negative value (in favour of the low economic group). Hence it can be stated that women with a low economic status are more sensitive to the improvement of the variables studied and receive greater benefit. However, as 42.3% of the women were from the low economic status group (while the same status was observed in 29.0% of the men), the women required greater economic support. In the explained portion, the constant value (intercept in the regression model) was significant and favoured the high economic group. This simply means that economic status per se has enabled the privileged group to have a better condition in terms of the outcome variable. Therefore, establishing economic equality will also be an effective measure in reducing the gap between the two groups.

In this study, age and education were recognised as the two most important factors in creating economic inequality. These factors have already been emphasised in other outcomes of the study.18 ,20 The important finding is that, although some studies have emphasised a relationship between blindness and economic status/poverty4 ,6 or between cataract and poverty,21 none has considered the role of economic inequality in near vision or presbyopia. Therefore, the results of this study could be useful from this particular perspective for policy making and for conducting other studies in the future.

High sample size and good response rate in sampling, conducting the examination accurately under constant supervision and employing the decomposition method, a method not frequently used, are the main strengths of this study. Nevertheless, one should note that the results have been obtained from a cross sectional study and the relationships identified are not necessarily causal. The results of the following phases of this cohort study will compensate for this deficiency. Shahroud city contains a representative population of Iran, and therefore our results can be generalised to the whole country.

To summarise, a significant gap is observed in PNVI between low and high economic status groups in the Iranian population. Age and education are two main factors affecting this gap, and any change in either of these two factors affects women with a low economic status to a greater degree.



  • Contributors MHE drafted the manuscript, contributed to the preparation of the study protocol, conceptualized and conducted all of the statistical analyses, and was the primary author of the article. HZ, RM, MS, HH and EJ contributed to the conceptualisation of the paper, the statistical analyses and critically revised the manuscript. AF conceived and designed the study, and contributed to the preparation of the study protocol, contributed to the conceptualisation of the paper and contributed to the statistical analyses, and critically revised the manuscript.

  • Funding This project was supported by Noor Ophthalmology Research Centre and Shahroud University of Medical Sciences.

  • Competing interests None.

  • Ethics approval The study was approved by the ethics committee of Shahroud University of Medical Sciences.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.