Article Text

Download PDFPDF

Ocular and systemic factors related to intraocular pressure in Japanese adults: the Tajimi study
  1. K Kawase1,
  2. A Tomidokoro2,
  3. M Araie2,
  4. A Iwase1,3,
  5. T Yamamoto1,
  6. Tajimi Study Group,
  7. Japan Glaucoma Society
  1. 1
    Department of Ophthalmology, Gifu University Graduate School of Medicine, Gifu, Japan
  2. 2
    Department of Ophthalmology, University of Tokyo Graduate School of Medicine, Tokyo, Japan
  3. 3
    Department of Ophthalmology, Tajimi Municipal Hospital, Tajimi, Japan
  1. Dr K Kawase, Department of Ophthalmology, Gifu University Graduate School of Medicine, Gifu, Japan, 1-1 Yanagido, Gifu-shi, 501-1194, Japan; mekawase{at}


Background: As intraocular pressure (IOP) and age are consistent risk factors of glaucoma, it is of special interest to know the association between IOP and possibly relating factors including age in Japan where a high prevalence of normal-tension glaucoma has been reported. The aim of this report was to evaluate the distribution of and factors related to applanation IOP in a population-based study in Japan.

Methods: A randomly sampled group of 3021 residents (response rate 78.1%) of Tajimi City, aged 40 years or older, underwent screening examinations including measurements of IOP with Goldmann applanation tonometry and central corneal thickness.

Results: Among right eyes without glaucoma, suspected glaucoma or other disorders which could affect correct IOP measurements, IOP averaged 14.6 (SD 2.7) and 14.5 (2.5) mm Hg in men and women, respectively, with no significant intergender difference (p = 0.342). Multiple regression analyses revealed that age was significantly negatively correlated with IOP (non-standardised beta (B) = −0.020/year, p = 0.0001). Higher body mass index (B = 0.14/BMI, p<0.0001), higher mean blood pressure (B = 0.022/mm Hg, p<0.0001), history of diabetes (p = 0.0019), thicker cornea (B = 0.014/μm, p<0.0001), higher myopia (B = 0.055/dioptres, p = 0.0043) and steeper corneal curvature (B = −0.72/mm, p = 0.0002) were also significantly correlated with higher IOP.

Conclusions: In an adult Japanese population, applanation IOP averaged 14.5 mm Hg and was negatively correlated with age after adjusting for other related factors. A positive correlation between IOP and myopia was found.

Statistics from

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.

Population-based studies have consistently shown that intraocular pressure (IOP) and age are the two major risk factors of glaucoma,14 making it important to investigate the distribution of IOP and its related factors, especially age, in various ethnic groups. From the standpoint of IOP, an adult Japanese population may be of special interest, because in this population the prevalence of open-angle glaucoma is relatively high (3.9%) but the mean IOP in eyes with open-angle glaucoma is relatively low (15.4 mm Hg)5 compared with other population-based studies.

Previous population-based studies from European or American countries reported that IOP increased with age,3 68 while those from Asia reported that IOP decreased with age.2 9 In these two Asian studies, however, IOP was measured with non-contact pneumatonometry, which may be less accurate and more likely to be affected by the central corneal thickness (CCT) than Goldmann applanation tonometry,10 and IOP was not corrected for the influence of CCT.

In the current report, we examined the distribution of IOP determined with Goldmann applanation tonometry and its correlation with other systemic and ocular factors, especially age, in Japanese adults who participated in an epidemiological glaucoma survey carried out in Tajimi City.5 The influence of factors reportedly correlating with IOP, such as blood pressure, body mass index (BMI), CCT, corneal curvature, refraction, a history of diabetes or smoking, was corrected by multiple regression analysis.


The subjects in the current study were identical to those in preceding reports on the Tajimi Study, which dealt with the prevalence of open-angle glaucoma. The details of the study have been reported previously5 and are summarised below. The investigation followed the tenets of the Declaration of Helsinki and the laws of Tajimi City regarding maintaining the privacy of patient data; the study protocol was approved by the local ethics committee of Tajimi City. Written informed consent was obtained from all participants.

Of 54165 inhabitants aged 40 years or older on 1 August 2000, in Tajimi City, based on the register of city inhabitants maintained by the municipal office, 4000 individuals were randomly selected with simple sampling using a table of random numbers without stratification and encouraged to participate in the epidemiological survey between September 2000 and October 2001. During the screening examination, the following ocular examinations were performed by well-trained ophthalmologists or technicians: objective measurement of refraction with an autokeratorefractometer (KP-8100PA, Topcon, Tokyo), best-corrected visual acuity (BCVA) using a Landolt ring chart (Inami, Tokyo) at a distance of 5 m, slit-lamp examination including assessment of the anterior chamber depth according to the van Herick method, CCT measurement using automated specular microscopy (SP-2000P, Topcon), IOP measurement using a Goldmann applanation tonometer (Inami), fundus photography (NW6S, Topcon), and visual-field testing using frequency-doubling technology perimetry with the C-20-1 screening protocol (Carl Zeiss Meditec, Dublin, CA). During the screening examination, the IOP was measured three times with a Goldmann applanation tonometer under topical anaesthesia, and the median value was adopted. The radius of the corneal curvature was determined as the mean of the radii of the steepest and flattest corneal meridians. If a subject satisfied one or more of the predetermined criteria to undergo a diagnostic examination,5 they were referred for that examination because of suspicion of ocular diseases including glaucoma or other diseases. A diagnosis of the eye in question was established by a panel of six expert ophthalmologists after discussion using all available documentation.

Statistical analysis

The relationship between a bilateral difference in IOP and systemic factors was assessed using the Spearman rank correlation analysis and linear multiple regression analysis in which the dependent variable was a bilateral difference in IOP, and the independent variables were systemic factors including age, BMI, mean blood pressure, a history of diabetes, and a history of smoking. The relationship between IOP and systemic or ocular factors was assessed using the Spearman rank correlation analysis and analysed further with multiple regression analysis in which the dependent variable was the IOP, and the independent variables were age, BMI, systolic and diastolic blood pressure, a history of diabetes, a history of smoking, CCT, the radius of corneal curvature, and the refractive error (spherical equivalent). A p value less than 0.05 was thought to be significant. The statistical analyses were performed using SPSS 14.0J for Windows (SPSS Japan, Tokyo).


Among the selected sample of 4000 subjects, 48 died, and 82 were non-residents or had moved from Tajimi City. A total of 3021 of the remaining eligible 3870 persons participated in the examinations, yielding a response rate of 78.1%. Of 6042 eyes (3021 participants), 860 eyes were excluded for various reasons that might have affected the results of IOP measurement, including the presence of glaucoma or the suspicion of glaucoma in at least one eye (460 eyes of 230 subjects) (table 1). As a consequence, 5182 eyes (2597 right and 2585 left eyes) were included in the current analysis.

Table 1 Reasons for exclusion from the current analysis of central corneal thickness in the participants of the Tajimi study

Among 2563 subjects of whom both eyes were eligible, the mean IOP values in the right and left eyes were 14.54 (2.57) mm Hg (95% CI 14.44 to 14.64) and 14.46 (2.57) mm Hg (95% CI 14.35 to 14.56), respectively, with significant bilateral difference (Wilcoxon signed rank test, p<0.001). The absolute difference in IOP between the right and left eyes averaged 0.40 (0.05) mm Hg (95% CI 0.38 to 0.43), which was significantly correlated with age (Spearman rank correlation coefficient (Rs) = 0.053, p = 0.008; partial correlation coefficient (PCC) = 0.049, p = 0.013) but not with sex (PCC = 0.005, p = 0.8), BMI (Rs = 0.012, p = 0.6; PCC = −0.006, p = 0.8), mean blood pressure (Rs = 0.045, p = 0.024; PCC = 0.024, p = 0.2), history of diabetes (PCC = −0.036, p = 0.074) or history of smoking (PCC = −0.005, p = 0.8). Because the IOP was highly correlated between right and left eyes (Rs = 0.956, p<0.001), and the results of left eyes were similar to those of right eyes, only the results from the right eyes were shown below.

The demographic data from the 2597 participants of whom at least the right eye was eligible are summarised in table 2. The IOP averaged 14.58 (2.61) mm Hg (95% CI 14.43 to 14.73) and 14.49 (2.53) mm Hg (95% CI 14.36 to 14.62) in men and women, respectively, with no significant intergender difference (p = 0.263, Mann–Whitney test) (fig 1).

Figure 1 Distribution of intraocular pressure in right eyes of 1154 men (black bar) and 1443 women (white bar).
Table 2 Demographic data for 2597 participants without glaucoma

Between IOP and age, a weak but significant correlation was found in both males (Rs = −0.067, p = 0.022) and females (Rs = −0.056, p = 0.032). As for the results of multiple regression analyses (R2 = 0.09), age was also negatively correlated with IOP (non-standardised beta (B) = −0.020/year, PCC = −0.077, p = 0.0001) (table 3). In addition, a higher BMI (Rs = 0.14, p<0.001; B = 0.084/BMI, PCC = 0.106, p<0.0001), higher mean blood pressure (Rs = 0.144, p<0.001; B = 0.022/mm Hg, p = 0.121, p<0.0001), history of diabetes (PCC = 0.062, p = 0.0019), thicker central cornea (Rs = 0.167, p<0.001; B = 0.014/μm, PCC = 0.175, p<0.0001), higher myopia (Rs = −0.090, p<0.001; B = −0.055/dioptre, PCC = −0.057, p = 0.0043) and a steeper corneal curvature (Rs = −0.055, p = 0.005; B = −0.72/mm, PCC = −0.074, p = 0.0002) were significantly correlated with higher IOP.

Table 3 Factors related to intraocular pressure in multiple regression analysis


In the current study, IOP measured by applanation tonometry averaged 14.5 mm Hg in a non-glaucomatous Japanese population aged 40 or older. Although this value was lower than that in normal-tension glaucoma patients (15.2 (2.4) mm Hg, right eyes, p = 0.006) in the Tajimi Study, it is notably higher than the mean IOP values previously reported for large samples of Japanese subjects, that is, 13.6 mm Hg by Shiose et al2 and 11.8 mm Hg by Nomura et al.11 In those studies, however, IOP was measured using a non-contact pneumatonometry, and the influence of CCT was not corrected. The subjects of the latter11 were office workers and their families, and therefore were not population-based samples, and the distribution of their age was apparently younger than that of the Tajimi Study. Although the study by Shiose et al2 was a population-based study performed in seven different cities widely distributed in Japan, the participation rate (ie, 50.5%) was not high enough. Therefore, the mean IOP value (14.5 mm Hg) in the current study should be a more appropriate estimate of the mean IOP measured using Goldmann tonometry in an adult Japanese population. When compared with the mean IOP measured by applanation tonometry in population-based studies from other countries, the mean IOP of 14.5 mm Hg in the current study is lower than those in the Baltimore Eye Survey (17.2 mm Hg in whites, 16.0 mm Hg in blacks),1 the Beaver Dam Eye Study (15.3 mm Hg in men, 15.5 mm Hg in women),6 and the Barbados Eye Study (17.8 mm Hg);7 is similar to those in the Rotterdam Study (14.6 mm Hg),12 the Egna-Neumarkt Study (15.1 mm Hg in men, 14.9 mm Hg in women),3 the Tehran Eye Study (14.5 mm Hg),8 the West Bengal Glaucoma Study (14 mm Hg)13 and a Bangladesh study (15 mm Hg);14 and is higher than that in a Thailand study (13.3 mm Hg)15 (table 4).

Table 4 Mean, SD and range of intraocular pressure (IOP) in population-based studies

Multiple regression analysis showed that the IOP was significantly associated with age, BMI, blood pressure, CCT and corneal curvature in men and women in this adult Japanese population. Older participants had a lower IOP; this trend is consistent with the results of a longitudinal observation for 10 consecutive years of 2987 young and middle-aged Japanese men in a corporate healthcare programme in which the IOP was measured by applanation tonometry.16 Nomura et al reported that IOP measured with non-contact tonometry decreased significantly with age in a large but non-randomised sample.11 The current report first confirmed the negative correlation between age and Goldmann applanation IOP in a Japanese adult population based on a randomly selected sample with a high participation rate. Many previous studies in the United States and Europe have shown that the IOP increased with age.3 6 7 In East Asia, IOP decreased with age among a population with an age of 65 and older in Taiwan9 and China.17 Thus, it is possible that the negative correlation between age and IOP is a common characteristic at least in people older than 65 years in East Asian countries. Further, the current study first demonstrated a negative correlation in a population including the younger generation (ie, 40 years or older).

In many previous population-based studies, a positive correlation between IOP and blood pressure was found in simple correlation analyses,12 multiple regression analyses,6 7 9 18 or longitudinal observations.19 20 This trend also was confirmed by multiple regression analysis in the current study. In general, blood pressure is higher in older individuals, which is still true in the population in the current study (Rs = 0.144, p<0.001). In Western populations, both IOP and blood pressure increase with increasing age, while in the current study, the older participants simultaneously had a higher blood pressure and lower IOP, suggesting the relationship between blood pressure and IOP independently of age. Although the true physiological mechanisms regarding a relationship between them is unclear, it has been speculated that increased blood pressure leads to increased ciliary artery pressure, increasing the ultrafiltration of the aqueous humour and thus increasing IOP.21

Positive correlations between IOP and BMI were reported in the Beaver Dam Eye Study6 and the Barbados Eye Study.7 The same trend also was seen in the current study. It has been suspected that in obese subjects elevated orbital pressure due to excess fat may lead to high episcleral venous pressure and decreased outflow facility.21 The presence of self-reported diabetes also was positively associated with IOP in the current study. A positive association between diabetes and IOP was reported in the Barbados Eye Study7 and the Rotterdam Study.22 Moreover, in the Baltimore Eye Survey, diabetes was reportedly associated with a higher IOP but not with the development of primary open-angle glaucoma.23 In the participants of the Tajimi Study, diabetes was not associated with the presence of primary open-angle glaucoma,4 which agrees with the Baltimore Eye Study.23

In the current study, a thicker cornea, more myopia and steeper corneal curvature were correlated with a higher IOP. A positive correlation between corneal thickness and IOP has been well known.18 24 To our knowledge, however, a relationship between IOP and refractive status has not been studied in population-based surveys except in the Beijing Eye Study,17 in which the IOP was measured by non-contact tonometry, and the CCT was not corrected. The current results confirmed a positive correlation between IOP measured by applanation tonometry and myopia, even after adjusting for the CCT and the other factors.

In summary, the mean IOP measured by applanation tonometry in a randomly sampled Japanese population 40 years of age or older was 14.5 (2.5) mm Hg. The IOP was significantly and negatively correlated with age after adjusting for other related factors. A positive correlation between IOP and myopia was found first as a population-based study. Further, higher BMI, higher blood pressure, a history of diabetes, steeper corneal curvature and a thicker cornea also were significantly correlated with a higher IOP value.



  • Funding: This work is supported by the Japan National Society for the Prevention of Blindness and the Japan Ophthalmologists Association.

  • Competing interests: None.

  • Ethics approval: The study protocol was approved by the local ethics committee of Tajimi City.

  • Patient consent: Obtained.