SEARCH

SEARCH BY CITATION

Keywords:

  • falls prevention;
  • frailty;
  • glaucoma;
  • physical activity;
  • visual field loss;
  • visual impairment

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Citation information: Black AA, Wood JM & Lovie-Kitchin JE. Inferior visual field reductions are associated with poorer functional status among older adults with glaucoma. Ophthalmic Physiol Opt 2011, 31, 283–291. doi: 10.1111/j.1475-1313.2010.00811.x

Abstract

Purpose:  To examine the relationship between visual impairment and functional status in a community-dwelling sample of older adults with glaucoma.

Methods:  This study included 74 community-dwelling older adults with open-angle glaucoma (aged 74 ± 6 years). Assessment of central vision included high-contrast visual acuity and Pelli-Robson contrast sensitivity. Binocular integrated visual fields were derived from merged monocular Humphrey Field Analyser visual field plots. Functional status outcome measures included physical performance tests (6-min walk test, timed up and go test and lower limb strength), a physical activity questionnaire (Physical Activity Scale for the Elderly) and an overall functional status score. Correlation and linear regression analyses, adjusting for age and gender, examined the association between visual impairment and functional status outcomes.

Results:  Greater levels of visual impairment were significantly associated with lower levels of functional status among community-dwelling older adults with glaucoma, independent of age and gender. Specifically, lower levels of visual function were associated with slower timed up and go performance, weaker lower limb strength, lower self-reported physical activity, and lower overall functional status scores. Of the components of vision examined, the inferior visual field and contrast factors were the strongest predictors of these functional outcomes, whereas the superior visual field factor was not related to functional status.

Conclusions:  Greater visual impairment, particularly in the inferior visual field and loss of contrast sensitivity, was associated with poorer functional status among older adults with glaucoma. The findings of this study highlight the potential links between visual impairment and the onset of functional decline. Interventions which promote physical activity among older adults with glaucoma may assist in preventing functional decline, frailty and falls, and improve overall health and well-being.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Preserving the health and well-being of older adults with visual impairment remains a challenge, particularly in light of our ageing population. The development of frailty poses a serious threat to all older adults, increasing their risk of falls, fractures, institutionalisation and mortality.1–5 Frailty is characterised by observable functional declines in the body, particularly weight loss, exhaustion, low energy expenditure, slowness and weakness,1,6,7 and has been linked to visual impairment in recent population studies.8–10

Glaucoma is a leading cause of irreversible visual impairment among older adults, leading to visual field loss, and in the later stages of the disease, to loss of central vision. As locomotion has been shown to be highly visually dependent,11,12 it is expected that mobility would become increasingly difficult with greater visual impairment. This has been shown in particular for visual field loss which results in slower walking speeds and increased number of obstacle contacts, both in general older populations13 and among those with glaucomatous visual impairment.14,15

Increasing difficulties with mobility are likely to influence participation in regular physical activity. Importantly, activity restriction among older adults has been shown to lead to reductions in physical function,16–18 and correlates with greater functional limitations and disability.19,20 As physical activity and exercise may postpone or reverse the effects of age-related loss of muscle mass and strength,21,22 early detection and prevention of frailty is an important factor in improving the health and well-being of older adults.

It is unclear, however, whether glaucomatous visual impairment is associated with reductions in physical activity or poorer functional status among older adults. In a case-control study which assessed falls and driving outcomes, basic mobility and functioning, measured using the timed up and go test, was significantly reduced for 48 glaucoma participants compared to 47 age-matched controls, but no differences were noted for self-reported physical activity levels.23 In a population study examining mobility performance, stair climbing speed among older adults with bilateral glaucoma was slower compared to those without glaucoma, but these differences were not statistically significant,14 although the findings are limited by the use of a non-standardised functional status measure. In our previous study, increased levels of visual field loss was associated with greater postural instability among older adults with glaucoma,24 which may also negatively influence participation in physical activity.

Despite the fact that frailty poses a serious threat to the health of older adults, there is also little known about the association between severity and location of visual field loss and functional status among older adults. Therefore, the aim of the current study was to examine the association between visual function, using a comprehensive battery of vision measures, and functional status in a community-dwelling sample of older adult with glaucoma.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Participants

Seventy four community-dwelling individuals aged 60 years and above and currently being treated for open-angle glaucoma were recruited from the clinical records of the Queensland University of Technology Optometry Clinic, private ophthalmology practices and local members of Glaucoma Australia. Participants were excluded if they had any significant ocular or visual pathway disease leading to visual field loss, other than glaucoma; any form of cataracts graded 3.0 or worse, defined by the Lens Opacities Classification System III;25 suffered from Parkinson’s Disease; history of dizziness or vestibular disease; used a walking aid; or had signs of cognitive impairment (Mini-Mental State Examination score <24 of 30).26 The research followed the tenets of the Declaration of Helsinki, and informed consent was obtained before participant assessment. The study was approved by the Queensland University of Technology Human Research Ethics Committee.

Demographic assessment

Data were collected on demographic information (age and gender) and medical information (medical history and current medication use). Fear of falling and the number of falls in the 12 months before participation in the study was determined by self-report.

Visual function assessment

Right and left visual acuity were measured with participants wearing their habitual distance refractive correction using a standard Bailey-Lovie high-contrast letter chart at a working distance of 6 m with a chart luminance of 160 cd m−2. Visual acuity was scored as the total number of letters read correctly, converted to logMAR units. Right and left letter contrast sensitivity was measured with habitual refractive correction using the Pelli-Robson letter chart at 1 m with a + 0.75 DS working distance correction in place,27 chart luminance of 83 cd m−2 and scored as the number of letters correctly identified.28 Better eye visual acuity and contrast sensitivity scores were used for the analyses.29,30

Visual fields were assessed with a computerized perimeter (Humphrey Field Analyzer; model HFA-II 750; Carl Zeiss Meditec Inc., Dublin, CA, USA). Monocular 24-2 Swedish Interactive Threshold Algorithm (SITA)-Standard threshold tests were performed by an experienced optometrist. A binocular mean deviation (MD) score was derived by merging the right and left fields to create an integrated visual field (IVF) extending 60° horizontally (IVF-60), based on the more sensitive of the two eyes at each visual field location.31,32 In addition, monocular 81-point, single intensity (24 dB) screening strategy tests were performed and merged to create a 96-point IVF extending 120° horizontally (IVF-120), based on the more sensitive of the two visual field locations in each eye, as outlined by Turano et al.13 The IVF-120 was scored as the total number of points missed. Points falling above and below the horizontal midline for the IVFs were used to determine the mean deviation scores (IVF-60) or points missed (IVF-120) for the superior and inferior field areas respectively.

Functional status assessment

The functional status tests were selected to represent a broad range of measures that are often associated with frailty, particularly muscle weakness, poor endurance, slowness and low physical activity.2,33 We selected standardised and validated tests that were suitable for use in a clinical setting, which included a number of performance-based and questionnaire measures. Participants wore their habitual distance refractive correction to complete the battery of functional status tests.

Lower limb muscle strength was measured using a spring gauge dynamometer, which measures the isometric strength of the knee extensor muscles (quadriceps), and has been shown to have good test-retest reliability (r = 0.75).34 While seated on a 65 cm high chair with hips and knees flexed to 90°, participants were instructed to extend their dominant leg smoothly and as forcefully as possible against the spring gauge strap which was attached above their ankle, generating maximum quadriceps force (in kg force). After one familiarisation attempt, three maximum voluntary contractions were completed and the maximum score was recorded. Higher scores reflect stronger lower limb musculature.

Mobility and physical functioning was examined with the 6 min walk test,35–37 which involves measuring the total distance (in metres) a participant can quickly walk along a level, well-lit, indoor corridor for a period of 6 min and has been shown to have excellent test-retest reliability (ICC = 0.97).38 The level of exertion is sub-maximal, as participants choose their own intensity level of exercise and were able to stop and rest during the test.37 Longer distances reflect better physical functioning.

Basic functional mobility and balance was assessed with the timed up and go test,39,40 which involves measuring the time taken to rise from a chair, walk 3 m at their usual walking pace, return, and sit down and has been shown to have good test-retest reliability in older adults (ICC = 0.97–0.99).38,39 Following one practice trial, the average time from two tests (in seconds) was recorded, where longer times reflect poorer functional mobility performance.

Self-reported physical activity level was assessed with the Physical Activity Scale for the Elderly (PASE).41 This is a self-administered 10-item questionnaire developed for older adults, where participation in activities in the 7 days prior to assessment is recorded, including leisure interests, sports, work, and household tasks. Participants completed this questionnaire during the testing session; good test-retest reliability among older adults has been demonstrated for self-administration of the PASE (r = 0.75).41 The PASE score was calculated by multiplying the time spent on each activity (hours/week) or participation in an activity (yes/no) by specific item weights, derived from measures of daily energy expenditure and self-reported physical activity, and summed over all activities, as outlined by Washburn et al.41,42 The questionnaire has also been shown to demonstrate moderate correlations with measures of health status, strength and balance (r = 0.25–0.42).41 Higher PASE scores reflect higher levels of self-reported physical activity.

An overall index of functional status was derived to capture participants’ overall physical functioning compared with that of the whole group. This approach has been used successfully in previous driving performance research.43 The score was formed using the converted Z-score values of the 6 min walk test, timed up and go test, lower limb strength and PASE scores; where necessary, the Z-scores were adjusted so more positive scores reflected better functional performance. The composite score was calculated as the mean of these Z-scores, with equal weight for all measures, and higher scores reflect better overall functional status.

Statistical analysis

The bivariate relationships between the visual function measures and functional status scores were examined using Pearson or Spearman correlation coefficients for the normally distributed and non-normally distributed variables, respectively. Correlations were adjusted for age and gender, as these variables were considered likely to be associated with both visual impairment and physical activity. p-values were not adjusted for multiple comparisons, as this approach can be considered overly conservative and can potentially mask important findings.44

Multivariate linear regression models were used to examine which specific vision factor/s were associated with each of the functional status outcomes, adjusting for age and gender. As the visual function measures were highly correlated in this cohort of glaucoma participants, the data were reduced by factor analysis to remove the influence of multicollinearity in the multivariate regression models.45 Factor analysis was performed on the contrast sensitivity and inferior and superior visual field variables, as they were significantly associated with the functional status outcomes at the bivariate level. The variables were submitted to principal components analysis using varimax rotation to derive three orthogonal statistically independent factors. The first factor loaded heavily on the superior visual field variables, the second factor loaded heavily on the inferior visual field variables, and the third factor loaded heavily on the contrast sensitivity variable. In total, the three factors accounted for 97.1% of the original variance of the data, with lower scores reflecting poorer visual function.

For each multivariate regression model, the coefficient estimates and 95% confidence intervals for each factor were calculated. The amount of variance explained by the three vision factors adjusted for age and gender were calculated by subtracting the coefficient of determination (R2) for the age and gender only models from the R2 for the full models. Residuals were evaluated to confirm the model assumptions of normality, linearity and homoscedasticity. All statistical analyses were performed using SPSS (version 16.0; SPSS, Chicago, IL, USA). p-values <0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The demographic, medical, visual function and functional status characteristics of the 74 participants are presented in Table 1. The mean age of participants was 74.2 ± 5.9 years, with slightly more males than females. The severity of glaucomatous visual impairment, based on visual field loss, ranged from early to advanced, with IVF-60 MD scores of −4.10 ± 6.28 dB (range −28.23 to 1.59) and IVF-120 points missed of 32 ± 21 (range 6–96). Participants using topical beta-blockers had similar levels of visual field loss (IVF-60 and IVF-120; p > 0.05) and functional status (all tests; p > 0.05) compared to those not using any topical beta-blocker glaucoma medications.

Table 1.   Characteristics of the study cohort (n = 74)
CharacteristicValue
  1. S.D. = standard deviation; IQR = inter-quartile range; IVF = integrated visual field.

Demographics
 Age, mean ± S.D. (range)74.2 ± 5.9 (62 to 90)
 Female, n (%)35 (47%)
Top 5 chronic medical conditions, n (%)
 Arthritis38 (51%)
 Hearing impairment29 (39%)
 Hypertension29 (39%)
 Heart disease22 (29%)
 History of cancer20 (27%)
Number of prescription medications, median (IQR)4 (2 to 7)
History of falls in the previous year, n (%)26 (35%)
Fear of falling reported, n (%)16 (22%)
Glaucoma medical history
 Number of glaucoma eye drops used, median (IQR)1 (1 to 2)
 Time since diagnosis, years, median (IQR)9 (4 to 15)
 History of previous glaucoma surgery, n (%)14 (19%)
Visual function measures
 Visual acuity in better-eye, logMAR, mean ± S.D. (range)0.06 ± 0.13 (−0.26 to 0.52)
 Contrast sensitivity in better-eye, logCS, mean ± S.D. (range)1.54 ± 0.17 (0.65 to 1.70)
 IVF-60 visual field, dB, mean ± S.D. (range)−4.10 ± 6.28 (1.59 to −28.23)
 IVF-120 visual field, points missed, mean ± S.D. (range)32 ± 21 (6 to 96)
Functional status measures
 Physical Activity Scale for Elderly (PASE) score, mean ± S.D. (range)128.7 ± 52.5 (37.9 to 301.3)
 6-min walk test, metres, mean ± S.D. (range)503 ± 69 (342 to 650)
 Timed-up and go test, seconds, mean ± S.D. (range)10.1 ± 2.0 (6.8 to 15.3)
 Lower limb strength, kg force, mean ± S.D. (range)19.4 ± 7.8 (6 to 44)
 Overall functional status score, mean ± S.D. (range)0.00 ± 0.71 (−1.75 to 1.62)

The age and gender adjusted correlations between the visual function measures and functional status outcomes are shown in Table 2. PASE scores were significantly associated with contrast sensitivity (r = 0.24) and all of the visual field measures (r=|0.23| to |0.31|). Timed up and go performance was associated with inferior IVF-120 (r = 0.23), while overall functional status score was only associated with IVF-60 (r = 0.25). In all cases, poorer visual function was associated with reduced performance on these measures. No significant associations were found for any of the vision measures and the 6-min walk test or lower limb strength.

Table 2.   Correlations coefficients of functional status outcomes and visual function measures, adjusted for age and gender (n = 74)
Visual Function MeasurePASE score6-min walk testTimed-up and go testLower limb strengthOverall functional status score
  1. Bold values indicate significant values.

  2. *p < 0.05; **p < 0.01.

  3. Pearson’s correlations.

  4. Spearman’s correlations.

  5. PASE = Physical Activity Scale for the Elderly; IVF = integrated visual field.

Visual Acuity, better-eye (logMAR)−0.150.10−0.130.04−0.15
Contrast Sensitivity, better-eye (logCS)0.24*−0.01−0.070.070.13
IVF-60, full field (dB)0.29*−0.03−0.060.060.15
IVF-60, inferior field (dB)0.31**0.08−0.170.110.25*
IVF-60, superior field (dB)0.25*−0.07−0.030.020.12
IVF-120, full field (points missed)0.28*−0.060.17−0.140.24*
IVF-120, inferior field (points missed)0.27*−0.140.23*−0.200.31**
IVF-120, superior field (points missed)0.25*0.030.09−0.06 −0.14

The loadings of the contrast sensitivity and visual field measures used to generate the three vision factors are shown in Table 3.

Table 3.   The loading of the visual function measures on the three factors (principal components factors analysis with varimax transformation; n = 74)
 Factor 1, ‘superior field’Factor 2, ‘inferior field’Factor 3, ‘contrast’
  1. Factor loadings >0.50 are in bold.

  2. The three rotated factors together represented 97.1% of the total variance.

  3. IVF = integrated visual field.

Contrast Sensitivity, better-eye (logCS)0.340.340.87
IVF-120, inferior field (points missed)−0.410.86−0.25
IVF-120, superior field (points missed)0.88−0.37−0.26
IVF-60, inferior field (dB)0.370.780.47
IVF-60, superior field (dB)0.830.370.36
% Variance explained (rotated solution)37.5%34.8%24.9%

The results of the multivariate linear regression analyses showed that poorer performance in the contrast factor was associated with lower PASE scores and overall functional status scores, adjusted for age and gender (Table 4). Furthermore, poorer performance in the inferior field factor was associated with slower timed-up and go scores, weaker lower limb strength and lower overall functional status scores. While poorer performance in the inferior field factor was associated with a longer 6-min walk distance, this failed to reach statistical significance. The superior field factor was not independently associated with any of the functional status outcomes. The vision factors explained the greatest amount of variance for the overall functional status scores (10.8%) and for the PASE scores (10.2%).

Table 4.   Multivariate regression findings for functional status outcomes on the vision factors, adjusted for age and gender (n = 74)
Vision factorPASE score6-min walk test (m)
Regression Coefficient (95% CI)p-ValueRegression Coefficient (95% CI)p-Value
Contrast factor11.5 (−0.1 to22.8)0.047−8.0 (5.3 to −21.3)0.24
Superior field factor−9.5 (2.2 to −21.2)0.1110.6 (24.3 to −3.1)0.13
Inferior field factor−9.2 (2.0 to −20.4)0.11−11.9 (1.2 to −25.0)0.076
Variance explained by vision factors (%)10.2% 6.1% 
Variance explained by the full model (%)14.7% 32.7% 
Vision factorTimed-up and go test (sec)Lower limb strength (kg force)Overall functional status score
Regression Coefficient (95% CI)p-ValueRegression Coefficient (95% CI)p-ValueRegression Coefficient (95% CI)p-Value
  1. Per unit decrease in vision factor.

  2. Wald chi-square test.

  3. Bold values indicate significant values; PASE = Physical Activity Scale for the Elderly; CI = confidence interval.

Contrast factor0.2 (0.6 to −0.1)0.23−1.1 (0.4 to −2.5)0.150.2 (−0.02 to0.3)0.028
Superior field factor−0.1 (0.3 to −0.5)0.540.3 (1.8 to −1.2)0.650.02 (0.16 to −0.11)0.76
Inferior field factor0.4 (0.1 to 0.8)0.0201.5 (−0.1 to2.9)0.0390.2 (−0.1 to0.3)0.004
Variance explained by vision factors (%)6.1% 5.5% 10.8% 
Variance explained by the full model (%)36.7%  36.6% 39.0%

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In this cross-sectional study, increased visual impairment was significantly associated with lower levels of functional status among community-dwelling older adults with glaucoma. Specifically, lower levels of visual function were associated with slower timed up and go performance, weaker lower limb strength, lower self-reported physical activity, and lower overall functional status scores. Of the visual factors examined, the contrast and inferior field factors were the strongest predictors of these outcomes, while the superior visual field factor was not related to any of the functional status outcomes. This is the first study to show a significant association between functional status and visual impairment among older adults with glaucoma, and the findings highlight the importance of vision in physical activity and physical functioning.

The mechanism of the relationship between glaucomatous visual impairment and functional status, albeit speculative, is that the glaucomatous visual loss may initiate restriction of physical activity due to difficulties in mobility, with subsequent declines in physical function due to inactivity. Previous studies have demonstrated significant reductions in mobility performance with more extensive visual field loss among older adults with glaucoma,14,15 other eye disease populations46,47 and general older populations.13 Experimental studies of younger normally sighted participants have also highlighted the negative impact of simulated visual field loss on mobility in virtual environments,48 and in gait behaviour, where more cautious gait strategies are adopted in the presence of field loss.49,50

Given that physical inactivity precedes reductions in physical functions,16,17 it was not surprising that stronger associations were found between the visual function measures and PASE scores, rather than with the performance-based measures. This was also evident in the multivariate models, where the vision factors explained nearly double the amount of variance in PASE scores (10.2%) than did the performance-based measures (5.5–6.1%).

Furthermore, this cohort of independent, community-dwelling adults had similar, if not higher, levels of physical function to other studies, which may have limited the strength of the correlations. Direct comparisons with other studies are difficult, however, due to variations in the age and health status of populations between studies. The mean PASE scores in this study were higher than other studies, including community-dwelling adults aged 65 years and over (pooled mean 106.4 ± 57.3)51 and glaucoma patients aged over 50 years (median 117, range 25–253).23 The mean 6-min walk distances were also longer than that recorded in non-institutionalized older adults older adults (pooled mean 406 ± 127 m),36 which is not surprising given that 30% of the participants in this study used a walking aid. Timed-up and go test scores were also comparable to research which included glaucoma patients (11 ± 3 s),23 and lower limb strength was similar to findings from population studies (pooled mean 23.8 ± 11.0 kg force).52

Our finding of an association between inferior field loss and functional status, rather than superior field loss, is consistent with studies which highlight the importance of this region of the visual field.12,13,53 When walking, individuals fixate approximately two steps ahead,54 and the inferior visual field contributes a major proportion of visual information used in lower limb movements, foot placement and obstacle detection.12 Studies involving adults with visual impairment have reported that greater loss in the central and inferior visual field regions results in reduced mobility performance.13,53 Furthermore, research has demonstrated that inferior visual field loss was associated with increased postural sway, reflecting greater instability, to a greater extent than did superior visual field loss.24

The contrast factor was also a significant predictor of a number of functional status outcomes, independent of the field factors, in our glaucoma group. This is consistent with previous research involving participants with other eye diseases and in a general aged population cohort, which demonstrated associations between contrast sensitivity loss and mobility performance13,46 and postural stability.45,55 It was not surprising that the association between visual acuity measures and functional status measures were weak, given the narrow range of visual acuity loss in this cohort, and hence low variability in these measures. Furthermore, visual acuity has not been shown to be strongly linked to mobility performance in previous studies involving heterogeneous visually impaired populations.13,46,47

The findings of this study are important, as the development of frailty increases the likelihood of serious adverse health outcomes, particularly falls, fractures, institutionalisation and mortality.1,2,5,33,56 Declines in physical function, functional limitations and frailty have been linked to central vision loss among older adults,8–10 and the present study provides additional evidence to support links between physical inactivity, functional declines and visual impairment among older adults with glaucoma.

Physical inactivity is likely to be an important link to the functional declines seen in this population. While physical activity and exercise hold great potential for improving and maintaining physical function among older adults,57 they may be challenging for visually impaired populations. Campbell et al.58 showed a tendency (which did not reach statistical significance) for a higher rate of falls among older adults with central vision loss who received an exercise program, compared to those not receiving the program. Their findings suggest that regular physical activity programs for visually impaired populations may be problematic, as these exercise programs had been effective in improving physical function and reducing falls among general community-dwelling older women in a previous study.59 A challenge for future research is to develop innovative physical activity programs which have the capacity to improve physical function and reduce frailty and falls for adults with visual impairment.

The strengths of the current study include the use of a battery of standardised visual function measures, including binocular integrated visual field measures, and the use of standardised measures of physical activity and functional performance designed for older adults. A potential limitation of this study, however, is the cross-sectional design, which precludes inferences about causality in the relationship between visual impairment and functional status. We cannot exclude the possibility that there was some recruitment bias towards more highly functioning participants, who were able to attend the research visits; this may have resulted in conservative estimates of the association between visual impairment and functional status. Furthermore, the regression coefficients need to be interpreted with caution, given the small sample size.

In summary, the present study provides important insights into the association between vision impairment and functional status among older adults with glaucoma, and identifies potential challenges in the prevention of functional decline. Visual field loss has been identified as an independent risk factor for falls and fractures among older adults,60–62 and innovative falls prevention programs for this population should include balance, strength and exercise training,63 to improve lower limb strength, postural stability and minimise frailty. A better understanding of the relationship between vision impairment and physical activity will help guide future strategies to promote and maintain the health and well-being of older adults with glaucoma.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank all of those people who participated in the study and Professor Beth Newman for her valuable assistance with study design and statistical advice. This project was funded by Queensland University of Technology and the Institute of Health and Biomedical Innovation and the first author received financial support through an Australian Postgraduate Award and a Queensland University of Technology Vice-Chancellor Scholarship.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References