The eye in focus: accommodation and presbyopia


Dr W Neil Charman
Faculty of Life Sciences
Moffat Building
University of Manchester
PO Box 88, Sackville Street
Manchester M60 1QD


Current understanding of the anatomy, function and performance of the accommodative system of the young, adult human eye is outlined. Most major current models of the accommodative mechanism are based on Helmholtz's original ideas but, despite of a growing volume of related research, uncertainty continues over the relative contributions made to the overall mechanism by different ocular structures. The changes with age are then discussed. Although the amplitude of accommodation decreases steadily from later childhood, the speed and accuracy of the system within the available amplitude are little impaired until the age of about 40, when the amplitude falls below that needed for normal near work. A review of the available evidence on age-related change in the lens, capsule, ciliary body and other relevant ocular structures confirms that geometric and viscoelastic lenticular changes play major roles in the progressive loss of accommodation. Other factors may also contribute in an, as yet, unquantified way and a full understanding of the origins of presbyopic change remains elusive.

It was, perhaps, inevitable that, following the early 17th Century development of optical instruments such as the telescope and simple and compound microscopes, together with their need for some method of adjusting focus to achieve clear imagery, attention should turn to the question of how the human eye manages to achieve optimal vision over a range of distances. Using his classic double-pinhole, Scheiner1,2 in 1619 demonstrated the existence of an active focusing mechanism and Descartes3 correctly speculated that this might involve changes in the shape and power of the crystalline lens. Surprisingly, nearly four centuries of further study have failed to clarify all the details of the accommodative mechanism and its changes with age. Although a broad but not unanimous consensus exists on the basic nature of the ocular changes involved, we still do not fully understand several important aspects, particularly the relative contribution of the age-related changes in each of the components of the system to the overall loss in accommodative amplitude with age. As will be evident from the article by Glasser4 in this issue, the need for improved understanding is pressing, in that the likelihood of success of different current and planned surgical approaches to the restoration or retention of some accommodation after the onset of presbyopia is governed by the correctness or otherwise of the assumptions made regarding the importance of the changes in specific parts of the accommodative system.


The accommodative plant

If we consider the optical components of the eye, it is clear that, in principle, focus could be changed actively by a variety of anatomical changes, ranging from a change in corneal curvature, through axial movement of the crystalline lens, to a change in axial length: passive mecha nisms might also result in a large depth-of-focus, rendering active focus change unnecessary (Figure 1). For example, given the dimensions and other parameters of the adult human eye, an accommodative response of four dioptres could be achieved by a reduction in corneal radius of about 0.7 mm or an increase in axial length of about 1.5 mm.5 Almost all of these and other possible mechanisms have had their advocates (see reviews6–12) but the elegant experiments of Young13 in 1801 resulted in most of these possibilities being eliminated as viable explanations of accommodative change. Numerous subsequent studies14,15 using such techniques as ophthalmophakometry and ultrasonography have confirmed Young's conclusion that any changes in corneal curvature or axial length are far too small to account for the observed changes in accommodation although, surprisingly, there is still debate on whether some minor changes in corneal curvature do occur.16–20 As is well known, there is now full agreement that changes in the power of the crystalline lens are overwhelmingly responsible for human accommodation and that there is little axial movement of the lens.

Figure 1.

Some possible mechanisms for focusing the eye

How are the required changes in the shape and power of the lens achieved in the young, adult, human eye? The classical view10,20–22 enunciated by Helmholtz6 and elaborated by Gullstrand,23 Fincham,24–26 Weale,27 Fisher28 and others is that both the lens and its capsule are elastic. When free of all constraint, the shape of the isolated lens is dictated by the balance between these elastic forces. With the capsule intact, the elastic forces that it provides, acting in opposition to the less powerful forces offered by the elasticity of the lens material itself, make the lens assume its most powerful, fully-accommodated form. Within the eye, the lens is supported by the meridionally-oriented anterior (or axial) zonular fibres. Rohen29 suggests that these are attached near the equator of the lens in three distinct sets: two of these are attached about 1.5 mm anterior and posterior to the lens equator and the third, finer set is attached along the equator itself. Glasser and Campbell30 disagree with the detail of this description and state that the attachment is essentially continuous across the equatorial region of the lens and along the internal edge of the ciliary processes, with many fibres crossing other fibres. Additional long posterior (or peripheral) holding fibres extend forward from the pars plana to pass through the valleys between the ciliary processes. Shorter tension fibres insert into the ciliary epithelium. The anterior zonular fibres place additional forces on the lens-capsule assembly. The ciliary body acts as a unit and contains a mixture of meridional, radial and circular muscle fibres.24,25 The circular fibres run round the free edge of the ciliary body, just behind the root of the iris. When accommodation is ‘relaxed’ for distance vision, the ciliary ring formed by the apex of the muscle is of relatively large diameter and the anterior zonular fibres are stretched by tension from the posterior pars plana fibres (Figure 2, top). The tension in the anterior zonular fibres exerts strong radial forces on the capsule, tending to stretch it. As a result, the balance of forces acting on the lens substance changes, causing the lens to flatten and attain the lower power required for distance vision.

Figure 2.

Schematic view of the mechanism of accommodation, as visualised by Rohen.29 In the unaccommodated eye (top) the ciliary muscle (CM) is relaxed. The anterior zonular fibres (AZ) are stretched by traction from the posterior (pars plana) zonular fibres (PZ). The resultant tension in the anterior fibres flattens the lens for distance vision. In the accommodated case (bottom), the forward and inward movement of the ciliary muscle (dashed curve) allows the tension fibre system to take up the tractional forces from the posterior zonular fibres and releases the tension in the anterior zonular fibres. The lens and its capsule can then take up their natural, more powerful, accommodated form (dashed curve).

Contraction of the ciliary body during accommodation, when the main muscle mass moves anteriorly and towards the axis, reduces the diameter of the ciliary ring. The tension fibre system is stretched, taking up the traction from the posterior zonular fibres. This allows the tension in the anterior zonule to relax and reduces the stretching forces acting on the lens capsule. The crystalline lens can then move towards its natural, more powerful form for near vision (Figure 2, lower). Other factors, particularly the elastic forces provided by the choroid also play a role.23 When accommodating, the stresses associated with the ciliary muscle contraction cause tension in the elastic choroid and may even cause anterior elongation of the retina.31–34 When unaccommodated, these tensional forces aid the relaxation of the ciliary muscle and thus help in the rapid restoration of the tension in the zonule, minimising the time taken to complete the response. Although some have suggested that the iris35 and vitreous body36–38 may also be important, the occurrence of accommodation in aniridic individuals26 and those lacking a vitreous body39,40 suggests that these factors can play only a subsidiary, rather than a major, role in the mechanism.

The changes in lenticular radius during accommodation are greater for the anterior surface of the lens,15,41–44 possibly because the tensional changes are greater in the anterior zonular fibres and the anterior capsule is thicker45 and exerts greater elastic forces on the lens substance. There is only a small posterior movement of the posterior lens surface, perhaps as a result of the resistance provided by the vitreous body, and a larger forward movement of the anterior surface. In combination, these result in an increase in axial thickness and a small forward movement of the centre of mass of the lens.12,46–48Figure 3 illustrates some typical data for these changes.

Figure 3.

(A) Changes in the surface radii of the crystalline lens as a function of accommodation in three young adult subjects (based on Rosales and co-workers46). (B) Variation with accommodation in the distances (mm), measured from the posterior surface of the cornea, of the poles of the anterior surface (red curve) and posterior surface (black curve) of the lens of a young adult emmetrope (based on Drexler and associates43). The difference between the two distances corresponds to the axial thickness of the lens.

The majority view is that the axial thickening of the lens in accommodation is accompanied by a decrease in its diameter.49,50 For example, Strenk and colleagues50 found that, with an eight dioptre accommodative stimulus, mean axial thickness in nine subjects aged between 22 and 30 years increased from about 3.6 to 4.0 mm, while the equatorial diameter decreased from 9.2 to 8.6 mm. Interestingly, Scheimpflug photography suggests that the axial thickening is almost entirely due to changes in the nuclear thickness of the lens, the anterior and posterior cortical thicknesses remaining almost constant as accommodation changes.41,46,51 It must be remembered that from the optical point of view the lens has a continuous gradient of refractive index, rather than being divided sharply into nuclear and cortical regions. It is generally assumed that the changes in the internal iso-index surfaces broadly follow those in the external lens surfaces but no reliable measurements have been made to confirm this. Undoubtedly the effects of the changes in the surface curvatures and those in the index gradients both make important contributions to the overall power changes in the lens during accommodation.52–55

Although the above explanation of accommodation has been largely accepted for many years, disagreements about its validity continue. Tscherning56 was an early dissenter, believing that contraction of the ciliary muscle would increase rather than decrease zonular tension and that as a result, during accommodation, the zonular forces would cause the lens to become flattened at its periphery but that the central region of the lens would bulge forward in the region of its axis to increase the power in the pupillary zone. Tscherning attributed this ‘anterior lenticonus’ to the differing mechanical properties of the lens cortex and nucleus. He assumed that the nucleus possessed both a greater curvature and resistance to mechanical forces, so that although the zonular forces flattened the peripheral regions of the lens, they served to increase the central curvature, which became dominated by the steeper curvature of the more-resistant nucleus. In this theory, during accommodation the lens diameter would be expected to increase and, if anything, the lens thickness would decrease slightly. More recently, a similar argument has been put forward by Schachar.57–59 He assumes that tensional forces on the lens equator increase during accommodation and that the lens assumes a similar shape to that suggested by Tscherning, although Schachar does not invoke the inhomogeneity of the lens in his explanation. He agrees with Rohen29 that there are three groups of anterior zonular fibres but suggests that the non-equatorial groups are responsible primarily for maintaining lens stability and it is the equatorial group in which the tension increases during accommodation. Arguments against these hypotheses include the fact that lens diameter appears to decrease during accommodation, rather than increase,26,49,50 while the axial thickness increases12,46–48 and the capsule seems to get slacker.26 Recent studies with optical coherence tomography60 suggest, too, that the anterior surface of the accommodated lens remains spherical rather than assuming the more complex form suggested by Tscherning56 and Schachar.57 The mathematical assumptions of the model of accommodation put forward by Schachar, Judge and Flavell61 have been disputed by Burd, Huang and Huang.62

Links with other aspects of near vision

Accommodation is often described as being linked with convergence and pupil constriction to form the ‘near triad’, implying that changes in each of the functions can potentially cause changes in each of the others. Although this is true in the case of accommodation and convergence, via the accommodative convergence/accommodation (AC/A) and convergence-induced accommodation/convergence (CA/C) ratios,63,64 it is not for changes in pupil diameter. Pupil constriction is a co-movement in near vision under independent control, which is not directly driven by other members of the near triad.65 While pupil constriction (accommodative miosis) generally occurs in response to a near stimulus, as observed at least as early as Scheiner in 1615,1 this is by no means always the case and there are wide individual variations in the extent of the constriction in subjects of the same age.66–71 Near miosis appears to be almost absent for most pre-teen subjects.68,69 Of course, any pupil constriction will affect the retinal image quality and tend to reduce the degrading effect of any dioptric error of focus or aberration. Therefore, it will tend to increase the subjective amplitude of accommodation, although not the objective amplitude.

Neural control

Contraction of the ciliary body is brought about predominantly by the activity of the parasympathetic fibres of the 3rd cranial nerve, which originate in the Edinger-Westphal nucleus.8,10,72,73 There has long been controversy over whether the sympathetic system plays any role in accommodation and this issue is still imperfectly resolved, although some weak supplementary inhibitory innervation is certainly present.74–77 The neural basis of the near response has been discussed by Mays and Gamlin78,79 and details of the involvement of different brain areas in accommodation are given by Judge and Flitcroft.80 Neural aspects will not be discussed further in this review.

Components of accommodation

Basing his ideas on Maddox's classification of the factors contributing to total convergence,81 Heath82 suggests that it is possible to distinguish several components of accommodation.

  • 1Reflex accommodation, the quasi-automatic adjustment of refractive state over perhaps a two-dioptre range to maintain a sharply-defined retinal image of the object of regard. (However, in numerous studies of accommodation apparently normal subjects are found who fail to accommodate to clear stimuli, which are modestly blurred.83,84)
  • 2Proximal or psychic accommodation, triggered by a knowledge (or a belief of a knowledge) of the object distance.85
  • 3Convergence accommodation, driven by fusion disparity vergence.63,64
  • 4Tonic accommodation, the slightly-myopic refractive state in the absence of a visual stimulus. This is manifested in the dark focus but is also found with an ‘empty’ or structureless visual field. It also corresponds to the preferred state of focus when using visual instruments (instrument myopia).86–88

It seems fair to include voluntary accommodation, as many individuals can train themselves to deliberately suppress or enhance the normal accommodative response under some circumstances.89–92 Heath82 considers ‘voluntary’ accommodation to be a form of proximal accommodation, contending that subjects are effectively ‘thinking of near’ to generate their accommodative response.

Retinal image quality during accommodation


Ideally, any focusing system should respond quickly and be capable of producing images of similar quality for objects over the full range of distances. How well does the eye of the young adult achieve this under photopic conditions? Changes in focus are not achieved instantaneously. There is a finite reaction time or latency of about 300 msec, followed by a further interval or response time of up to about 1000 msec during which the accommodation is changing before stabilising at its new level.93–98 Examples are shown in Figure 4.

Figure 4.

Examples of monocular accommodative responses to abrupt step changes in the vergence of an accommodative stimulus.83 Note the time taken (latency or reaction time) before the response starts and the time taken to complete the response (response time). Fluctuations in accommodation can be seen. The lower traces merely show the times at which the stimulus changes occurred. The actual stimulus changes are between (A) 2.38 and 1.33 D and (B) 1.33 and 2.38 D.

If the level of accommodation to a constant stimulus is examined with good time resolution, small fluctuations in accommodation are observed, having an amplitude of about 0.1 D and a frequency spectrum extending to a few hertz99 (Figure 4). It is thought that the low-frequency components of these fluctuations may have a role in helping to maintain the steady-state response.100,101

When the mean steady-state response level achieved is compared with the stimulus level over the available range of accommodation, a plot of the form of Figure 5A is usually found.20,21 There is over-accommodation (‘accommodative lead’) for distant objects and under-accommodation (‘accommodative lag’) for near objects. At distance, the lead is usually found because most conventional least negative/most positive refractive techniques leave the true refraction of the eye slightly myopic at the conventional testing distance of six metres, relying on the individual's depth-of-focus to maintain optimal acuity. The slope of the quasi-linear central portion of the response/stimulus curve varies with the individual (particularly with age and visual acuity) and with the nature of the target under observation (its form, contrast and colour) and other conditions (particularly luminance, see below).20,21,102 A key feature is that, under many conditions, the retinal image shows a significant error of focus of about 0.25 D or more, even within the individual's amplitude of accommodation. These errors in focus are usually the major causes of retinal image degradation in the accommodated eye, being much more important in their effects than higher-order aberrations. Johnson103 was able to show that, because of the changing leads and lags of the type shown in Figure 4A, visual acuity was optimal at the slightly-myopic level of accommodation, at which minimum error of focus occurred. This level corresponds to Heath's tonic level of accommodation.82 It is interesting to note that two recent studies104,105 have shown that surprisingly high levels of blur (up to around 1.5 D when reading text, well in excess of those that are just detectable) may be tolerated without complaint by many people. Large accommodative lags have previously been found in children, particularly myopes.106

Figure 5.

(A) Schematic appearance of a typical accommodation/response stimulus curve under photopic conditions. The black portion of the curve represents the range, over which the target is seen without noticeable blur: the lighter grey portions indicate where the lags or leads in accommodation exceed the depth-of-focus and the target appears blurred. OA is the objective amplitude of accommodation and SA the subjective amplitude. The ‘ideal’ line corresponds to equal response and stimulus values. (B) Typical changes in the lower part of the response/stimulus curve with target luminance (based on Johnson103). Target luminances are as indicated and the independently-measured dark focus for the same subject is 2.0 D, corresponding closely to the cross-over point of the response curves. (C) Frequency distribution of tonic accommodation as estimated by the dark focus: the values shown represent the differences between autorefractor measurements of refraction under photopic conditions and in darkness.116

Dynamic and static accommodative responses are normally similar in both eyes, suggesting that they have a common neural origin.107–109 There is some disagreement about whether appropriately unequal steady-state responses may occur when the accommodative demands to the two eyes differ, for example, when a close object is placed to one side of the midline, with some authors finding significant differences in response (up to 1.0 D),110–112 while others fail to do so.113,114

Active accommodation appears to depend on cone activity.115 If the luminance level is progressively lowered, the slope of the response/stimulus curve falls steadily until, when only the rods are active, the accommodative system ceases to function103 (Figure 5B). Accommodation then assumes the constant, slightly-myopic ‘tonic’ level (the ‘dark focus’) that, in the absence of a stimulus, would be expected from Heath's classification of the components of accommodation,103,115 as only the tonic component is then active. The tonic level varies with the individual. Figure 5C shows typical data for young adult subjects.116

Changes in aberrations

Accommodation changes both surface curvatures of the crystalline lens and its internal gradients of refractive index. It is not surprising to find changes in ocular aberrations as the level of accommodation changes. As discussed above, it is likely that under most circumstances errors in spherical focus will be the largest contributors to retinal image blur in the accommodated eye. Other second- and higher-order aberrations may also play a role, their relative importance increasing under conditions where the lags and leads in focus are small.

The most obvious contributor is accommodation-dependent change in second-order astigmatism. Early workers speculated that, if an eye possessed uncorrected astigmatism at distance, different levels of accommodation might occur deliberately in different meridians to correct the astigmatism at near. No evidence for this effect has ever been found.117–121 Nevertheless, subject-dependent changes in ocular astigmatism do occur and recent studies suggest that, on average, second-order astigmatism may change in the with-the rule direction by about 0.04 D for each dioptre of accommodation response.122,123 If it is accepted that the zonule slackens during accommodation, this could be due to a change in lens tilt.

Initial attempts to explore the changes in higher-order monochromatic aberrations that accompany accommodation correctly identified a trend for the spherical aberration to move in the negative direction, that is, to move from under- to over-correction and become less positive/more negative.124–127 Early workers were hampered by the limitations of their equipment, in particular by its inability to make simultaneous measurements of ray or wave aberrations over the whole pupil. Recent developments in aberrometry128,129 have overcome this difficulty and a number of groups have explored the changes in large groups of subjects.18,122,130–133 While inevitably there are minor differences in their results, mainly due to the use of different subject groups and measurement conditions, there is broad unanimity in the general trends of the data for spherical aberration. All agree that, for young adults, this moves in the negative direction as accommodation is increased. Coma also changes but different groups find different directions and magnitudes of change. Changes on other higher-order aberrations appear to be small and to vary between individuals.122,132 The changes in spherical aberration may be relevant to the changing second-order focus errors (lags of accommodation) with stimulus level.132

As the overall power of the eye increases during accommodation, it would be expected that the longitudinal chromatic aberration of the eye would increase in an approximately similar way,134–136 by around two per cent per dioptre of accommodative response. This appears to occur in practice137 but the effect is of minor importance. It is possible that, due to changes in lens tilt and centration and possibly pupil centration, accommodation may have larger effects on the transverse chromatic aberration and polychromatic image quality, as recorded on the visual axis but these effects do not appear to have been investigated.

The presence of aberrations may aid accommodation. Defocus blur alone does not provide cues to the direction of the response required to bring an image into focus.138–140 There has long been speculation that aberrations might aid the accommodation system of at least some individuals by providing odd-error cues to the appropriate direction of response change, when the eye is presented with an out-of-focus target. It appears that this is true for chromatic aberration141–146 and recent studies also give some support to this suggestion in the case of monochromatic aberrations,75,147,148 although Lopez-Gil and his colleagues149 found that contact-lens induced changes in third-order aberrations had little impact on the dynamics of accommodation.

Models of the accommodative mechanism

A variety of static and dynamic models has been developed, based on data obtained from the numerous experimental studies of different aspects of the accommodative system. Dynamic biomechanical models that include the viscoelastic characteristics of the lens, zonule and choroid have been produced by several authors.96,150–153 Other models suggest that the elastic anisotropy of the lens and support from the vitreous may play important roles.154,155Figure 6 shows the basis of the dynamic model of Beers and van der Heijde,96 which relates directly to the anatomy and physiology of the accommodative system. The elastic choroid and posterior (peripheral) and anterior (axial) zonules are represented by springs: the lens is represented by a spring corresponding to the capsule and a parallel dashpot to take account of the cytoplasmic movement of the lens fibres during accommodation. When appropriate time constants and other parameters are included, the model gives useful predictions of dynamic accommodative behaviour. Several attempts have been made to model the accommodating lens numerically, using finite element methods.61,62,156–158 Finally, substantial effort has been devoted to the development of static and dynamic control system models, with the more elaborate including the full near triad and a number of feedback loops and cross-links.85,159-167 While all these models have been successful in some respects and have proved helpful to the understanding of accommodative behaviour in clinically unusual cases,168–170 all require further refinement before they can fully simulate observed accommodative behaviour.

Figure 6.

The dynamic biomechanical model of Beers and van der Heijde.96 The elements of the model are shown superimposed on the anatomy of the components of the accommodative system (after Beers and van der Heijde96). The choroidal spring is anchored at the posterior pole and the lens spring and dashpot system at the lens centre. The model has been applied to the young adult eye96 and to age-related changes in accommodation.153


Amplitude of accommodation

The progressive decline in the amplitude of accommodation that occurs with age was recognised at an early date, although for many years there was some confusion between the effects of hypermetropia and presbyopia.7 Indeed, it is interesting to note that Donders in his magisterial work Accommodation and Refraction of the Eye was still using the term ‘hypermetropia acquisita’ as a synonym for presbyopia.171 Although there appears to be no general agreement on how to define the onset of presbyopia, it is often accepted that this term should be applied when the subjective amplitude of accommodation falls below three dioptres.172

Over the last century or so, numerous studies have been made of the way in which amplitude (measured under photopic conditions) falls with age. Subjective amplitudes are slightly higher than objective amplitudes, due to the inclusion of depth-of-focus effects.173–175 Longitudinal studies suggest that individual objective amplitudes fall almost linearly with age to reach zero at about the age of 50,174,175 although averaged data may show a reduced rate of decline as presbyopia is approached due to the averaging of individual linear declines with different intercepts.176 There is some suggestion that the intercept may vary with latitude or temperature at the individual's place of residence177–179 but, in general, the striking feature is that the changes with age vary very little between individuals and, unlike many aspects of human physiology, do not appear to have been affected by the changes in nutrition and lifestyle of the modern age. Current amplitude measurements are very similar to those recorded by Donders in the 19th Century171 (although it must be remembered that Donder's results were obtained with maximal convergence and were referred to the nodal point of the eye, seven millimetres behind the cornea, rather than the spectacle plane180). Typical mean, transverse, monocular data181–184 as found by several authors are shown in Figure 7.

Figure 7.

Results of transverse studies of the mean monocular subjective amplitude as a function of age. (Based on Duane,181 Coates,182 Turner,183 Ayshire Study Circle.184)

While there are some differences between the data of different investigators, presumably reflecting differences in the methodology used to record the amplitudes and in the subject populations,185–187 the striking feature of the amplitude curve is, of course, the fact that amplitude is already reducing well before maturity is reached, in contrast with other aspects of visual performance such as visual acuity, which only start to show decline after the age of about 50.188,189 It is, of course, possible to view these changes more positively and to say that, in youth, we are over-endowed with accommodative ability. For example, if it is assumed that a near point of, say, 0.25 m is adequate for most tasks, then problems will start to arise only at the age of about 40 years. This corresponds to the expectation of life in many primitive societies, so that presbyopia may not have been a problem from the evolutionary point of view.

Changes in other aspects of accommodation with age

Although changes in amplitude are the most familiar, it is also of interest to ask: what are the changes in the speed and accuracy of the responses with age? as these may improve our understanding of the causes of presbyopia.

Considering first the temporal characteristics of the response, if measurements of step responses are made within the available range of accommodation, the striking finding is that, although amplitudes of subjective accommodation change markedly between, say, 20 and 40 years, reaction and response times to step stimuli show at most only modest changes. There is some disagreement between different authors regarding reaction time changes. Ciuffreda and Mordi190 found an increase in latency in both accommodation and disaccommodation, from about 350 to 400 msec between the ages of 20 and 45 years, whereas Kasthurirangan and Glasser191 suggested that only the latency for disaccommodation increased (from about 200 to 300 msec over the same age range): Heron and his group83,192–194 found little difference in the two reaction times. Several studies83,191–194 were unable to find convincing evidence for any marked age-dependent change in the time course of either positive or negative responses. Kasthurirangan and Glasser191 found that constants for accommodation increased with age, although those for disaccommodation did not. If responses to stimuli that vary sinusoidally with time are considered, it appears that the ability to respond to the stimulus is well preserved up to the age of about 40 years (Figure 8), the cut-off frequency at which accommodation can no longer respond to the changing stimulus being about two hertz.83,192–194 There is a slow decline in the amplitude of the response to a fixed stimulus change and an increasing phase lag, the latter corresponding to a fixed time delay, which increases with age.

Figure 8.

Accommodation responses of a younger (21 years, A, B) and older (41 years, C, D) subject to accommodation stimuli, which vary sinusoidally with time at either 0.1 or 0.6 Hz, between 2.38 and 1.33 D. The curves representing the stimuli have the correct phase but their amplitude and mean level have been reduced to avoid overlap with the response traces. The older subject can still follow the higher-frequency stimulus changes but with reduced amplitude and a greater phase lag.83

The steady-state response/stimulus curve also maintains its form well up to the age of 40 years. The linear portion of the curve shows only modest changes in slope, although its extent gradually reduces with the reduction in the amplitude of accommodation. Beyond the age of 40 years or so, the slope starts to decrease markedly, probably because the system attempts to optimise the use of depth-of-focus and the small available objective amplitude to obtain adequate vision over as wide a range of object distances as possible.71,175,195,196 Some typical data illustrating these general trends are shown in Figure 9.

Figure 9.

Examples of steady-state response/stimulus curves over the stimulus range 0.5 to 5.0 D, for subjects within the age groups indicated. Each curve represents a single subject (after Kalsi, Heron and Charman196).

The dependence of aberration on accommodation also changes with age,133 almost certainly because of the changes occurring in the crystalline lens. Measured at constant pupil diameter, aberrations tend to increase. Comparisons of the visual impact of aberrations for individuals of different ages are complicated by the gradual reduction in pupil diameter that occurs with age. Calver, Cox and Elliott197 suggest that the effect of this is to make the modulation transfer function measured at constant illumination almost independent of age. For younger subjects spherical aberration tends to move in the negative direction as accommodation is increased. For those less than 20 years it is near zero at levels of accommodation of around 0.5 D, the corresponding value rising to around two to three dioptres in subjects between 20 and 40 years. Subjects aged over 40 years appear to experience an increase in positive spherical aberration, as they exercise their limited range of accommodation.133

Early studies of the links between accommodation and convergence as a function of age produced mixed results198,199 but later studies are more consistent. Bruce, Atchison and Bhoola200 found that, for subjects aged between 15 and 45 years, the response AC/A ratio increased with age and the response CA/C ratio decreased. Broadly compatible results for the same ratios have been found by other authors, although extending over slightly different age ranges.201,202 Ciuffreda, Rosenfield and Chen203 found a similar marked increase in response AC/A between the ages of 20 and 45 years but there was no significant change when they restricted their analysis to subjects between 20 and 35 years of age. Overall, the evidence favours an increase in response AC/A through pre-presbyopic adult life, a characteristic of some importance in relation to theories of presbyopia, as it suggests that increased innervation is required in the older eye to produce the same accommodation response (see below). Accommodative miosis varies widely between subjects of the same age and does not appear to change systematically between the ages of 20 and 40 years.68–71

Ageing of the components of the accommodative system


As is well known, the lens increases in thickness and weight through the addition of new lens fibres throughout life. The new fibres originate in the equatorial region and grow towards the axis across the anterior and posterior surfaces of the lens under the capsule.72,204 Numerous studies have been made of the in vitro age-dependence of the lens parameters (see reviews30,205–207). Table 1 gives recent data for some of the changes, deduced from measurements of 37 lenses over the age range 20 to 99 years.207 As the in vitro lens is free of all zonular forces and on the basis of classical theories of accommodation, it would be expected that the isolated lens would assume its fully accommodated form.208 It can be seen in Table 1 that all of the listed parameters increase with age, with the exception of the posterior radius of curvature, which remains constant. Other authors30,209 suggest that, rather than the changes in anterior radius of curvature being linear with age, there may be a discontinuity at the age of 65 years with a linear decrease thereafter.

Table 1. Regression line fits to the in vitro changes in some lens parameters with age: x is the age in years (from Rosen and associates207 for lenses in the age range 20–99 years)
ParameterRegression equation
Lens weight (mg)W = 1.45x + 164
Total sagittal thickness (mm)T = 0.0123x + 3.97
Equatorial diameter (mm)D = 0.0138x + 8.7
Anterior radius of curvature (mm)Ra= 0.046x + 7.5
Posterior radius of curvature (mm)Rp = 5.5

Accompanying these changes in the external characteristics of the in vitro lens are changes in the internal gradients of refractive index. As the lens ages and its axial thickness increases, the lens becomes optically more homogeneous, with the zone over which the index is changing being increasingly confined to the cortical layers of the lens, while the index over most of the interior remaining essentially constant.210,211 Refractive indices at the lens centre and at the outer surface of the cortex are constant with age, with values of 1.418 ± 0.075 and 1.371 ± 0.004, respectively.211 These changes in overall shape and index affect the lenticular aberrations, particularly spherical aberration, which becomes increasingly positive as the lens ages.205 The changes in the index gradients of the in vitro lens as a function of age, together with the expected associated changes in lens power and spherical aberration, have recently been modelled analytically.212,213

The mechanical characteristics of the lens and its capsule have also been explored in vitro. In a series of experiments, Fisher214–217 determined the elasticity and other mechanical characteristics of the lens and its capsule as a function of age. Interestingly, Young's modulus of elasticity for the capsule was found to decline steadily with age by a factor of about two between 20 and 60 years, although the capsular thickness increased and the lens stiffness increased (Figure 10A). Glasser and Campbell30 also found that the resistance to deformation increased with age and made the important observation that, when the lens capsule was removed, there was a marked decrease in power of younger lenses but little change in older lenses, suggesting that the capsular forces had little effect on the lens substance of the isolated lens after the age of about 50 years. There is little change in the water content of the lens with age and it has been suggested that the changes in the mechanical characteristics of the lens occur as a result of increased adhesion and compaction of the nuclear fibres.218 Pau and Kranz219 found a distinct hardening of the nucleus with age. It should be added that the assumptions underlying Fisher's interpretation of his lens measurements, in which the derivation of stiffness parameters involved spinning the lens about its axis,216 have recently been challenged by Burd, Wilde and Judge.220 In particular, they cast doubt on Fisher's conclusion that the cortex is significantly stiffer than the nucleus, notably in middle age.

Figure 10.

(A) Changes in Young's modulus of elasticity for the anterior capsule as a function of age (after Fisher215). (B) Changes with age in the maximum force of contraction exerted by the entire ciliary muscle (after Fisher230). Vertical lines indicate standard deviations within each decade.

In vivo, the dimensional parameters of the lens have been measured as a function of both age and accommodation.41–43,46,50,51,208,221,222 As an example of such data, Table 2 shows regression results from a recent study222 in which measurements were made both by Scheimpflug photography and by high-resolution magnetic resonance imaging of the lens when accommodation was relaxed. The diameter of the unaccommodated (relaxed) lens (around nine millimetres) does not change with age but its reduction in diameter when presented with a strong accommodative stimulus (eight dioptres binocularly) falls from a value of about one millimetre at the age of 20 years to zero at around 50 years.51

Table 2. Regression line data for the age changes in in vivo crystalline lenses with relaxed accommodation: x is the age in years (from Koretz and co-workers,222 subjects between 18-50 years)
Sagittal thickness (mm)T = 0.0194x + 3.088T = 0.0193x + 2.944
Anterior radius of curvature (mm)Ra = -0.0759x + 13.949Ra = -0.0828x + 13.443
Posterior radius of curvature (mm)Rp = -0.0106x + 6.436Rp = +0.0078x + 5.368

Again, we can see that the changes with age in the posterior radius of curvature are much smaller than those for the anterior surface and, indeed, that the two methods differ in the sign of the small gradient of the age changes for the posterior surface. The anterior radius of curvature of the relaxed eye diminishes with age. Naively, it would be expected that this would lead to an increase in lens power and in turn to the development of myopia. It has been suggested that the reduced importance of the contribution that the gradient of refractive index makes to the overall power of the older lens results in the total power remaining almost independent of age,54,223–225 although Glasser and Campbell205 suggest from their in vitro experiments that stretched (that is, ‘unaccommodated’) lenses decrease in power with age.


The changes in the ciliary muscle with age have been described by several authors.226,227 Several morphological changes occur. In particular, there is an increase in the amount of connective tissue: the muscle also becomes shorter and wider, while the apical edge moves forward. In the relaxed eye, the ciliary ring diameter and the circumlental space between the ciliary ring and the lens equator decrease with age, while the lens diameter remains almost constant.51,228 In agreement with earlier studies based on impedance cyclography,229 Pardue and Sivak227 find that the muscle retains its ability to contract throughout the lifespan. Interestingly, the impedance cyclographic data suggest that the speed of relaxation of the muscle during disaccommodation is less in older than in younger subjects.229

Fisher230,231 suggested that the maximum force of contraction of the entire ciliary muscle increases with age to reach a peak at the age of about 50 and decline slightly in later years (Figure 10B). The ciliary ring contracts by about 0.8 mm in radius during maximum accommodation; this change in radius is essentially constant between the ages of 15 and 45 years. Fisher found that the force of contraction of the ciliary muscle was proportional to the square of the resultant accommodative change, rather than being linearly related.


It is reasonable to suppose that the direction and strength of the forces applied by the zonule to the capsule will change, if the positions of the attachments of the fibres change. Farnsworth and Shyne232 showed that the distance between the anterior zonular insertion and the lens equator increased with age and speculated that this might have a role in the development of presbyopia. Their results have since been extended by Sakabe and colleagues,233 who found that the zonular-free zone of the anterior surface of the lens decreased in diameter. Closely related to these findings was Brown's observation234 that the shape of the lens equator changes with age and the course of the zonular fibres becomes more divergent. Fisher231 found that the extensile properties of the zonular fibres do not change over the age range 15 to 45 years.


Few studies appear to have been made of ageing of the choroidal part of the accommodative apparatus. The apparent retinal stretch with accommodative effort31–34 was observed in a presbyopic 60-year-old eye,32 suggesting that the ciliary muscle/choroid part of the accommodative complex was still active. On the other hand there is some evidence that the elastic tendons that form the posterior attachment of the ciliary muscle in rhesus monkeys and are continuous with the elastic laminae of Bruch's membrane may thicken and become more rigid with age, limiting the anterior-inward movement of the ciliary muscle during accommodation.235,236


As discussed earlier, some theories of accommodation suggest a role for the vitreous,36–38 particularly in acting as a support for the rear surface of the lens, although accommodation occurs in eyes lacking a vitreous.40,41 The vitreous becomes more liquid with age and hence might be less capable of fulfilling a supportive role,237 although it is difficult to see how this process could lead to a linear decline in amplitude with prepresbyopic age. Weale238 suggests that the insertion of the iris root may be of importance and there is some evidence that accommodation is lower in iridectomised monkeys.239 Obvious changes occur in the iris geometry as a result of the age-dependent reduction in the depth of the anterior chamber, itself a result of increasing lenticular thickness.

Explanations of presbyopia

In the light of the various studies of the age changes in the structure and performance of the accommodative system, a surprisingly large range of theories has been developed. These are well summarised in several recent reviews.10,12,238,240–242 Only those theories that are more widely supported and are relevant to possible attempts to restore accommodation are briefly outlined here.


While capsular forces show only modest changes, it is clear that the ageing lens becomes harder to deform. Thus, even if the ciliary muscle retains its power and the capsule most of its elasticity, collectively they cannot reshape the lens appropriately. On this basis, if the problem of the mechanical resistance of the lens substance could be overcome, for example by refilling the capsule with some suitable man-made medium,4 accommodation could be restored.

In fact, two major variants of the lenticular theory are often described.10,20,21,150,240 These are usually named the Hess-Gullstrand and Duane-Fincham theories, although as usually outlined, they owe as much to the interpretation of later authors as to the ideas of their eponymous originators. In the simplest statement of the former theory, it is assumed that throughout life a constant amount of ciliary muscle contraction is required for each dioptre of accommodative change. Thus, as the amplitude of accommodation diminishes due to lenticular changes, an increasing proportion of potential ciliary muscle contraction will be ‘latent’ in the fully-accommodated eye, in the sense that further contraction will produce no accommodation change. In contrast, the Duane-Fincham theory assumes that as the eye ages, the force required to produce a given change in accommodation increases and that the ciliary muscle will be maximally contracted when the near point is reached. Duane181,243 thought that the ciliary muscle would weaken with age but, in agreement with Fisher's230 later findings, Fincham25,26 did not. It is often suggested10,20,21,150,240 that within the available range of accommodation, the Hess-Gullstrand theory predicts that the slope of the response/stimulus curve is constant with age, whereas the Duane-Fincham slope declines with age so that lags in accommodation at any stimulus level become progressively larger though life. This ignores the fact that to yield clear vision when accommodating, the system must maintain focus within the depth-of-focus, so that similar slopes might be expected on the basis of either theory.175,196

Fisher's experimental studies230 appear to rule out the Hess-Gullstrand theory in its simplest form, as he finds that the relationship between ciliary force and accommodative response is not linear. Moreover, there is ample evidence that greater force is required to produce a given change in accommodation as the eye ages.30,230 Changes such as the observed increase in response AC/A ratio with age199–202 also appear to refute the Hess-Gullstrand theory. Studies like those of Fisher214,216,217 and Glasser and Campbell30,205 suggest that, as the eye ages, greater force will need to be exerted on the lens and capsule to achieve the same change in accommodation, thus supporting the basic Duane-Fincham model. Changes in the dynamic response characteristics as a function of age have been interpreted to indicate that changes in the elastic properties of the lens are the prime cause of presbyopia.153

It may be that any final model will have to incorporate elements of both the Hess-Gullstrand and Duane-Fincham models. Thus while, with age, greater ciliary force may be required to produce the same response, it may also be true that some reserve of ciliary force remains untapped when the lens has reached its fully-accommodated state, as must indeed be the case when the eye becomes fully presbyopic and no change in lenticular power occurs.


Duane181,243 proposed that accommodative loss occurred because of the progressive weakening of the ciliary muscle. As noted earlier, substantial evidence230 contradicts this possibility. A related possibility is that although the ciliary body maintains its ability to contract, the elastic forces available within the ciliary body and choroid, which oppose this contraction and provide the tension required to aid the relaxation of accommodation, decline.244 This would leave the lens in a progressively more accommodated state, with greater surface curvature at zero accommodation, as observed.


These are based on the age-related changes in the shape and size of the lens, rather than on changes in its mechanical characteristics. As noted earlier, the lens curvature and thickness increase and the distance of the insertions of the zonular fibres into the capsule from the equator is increased.232,233 Koretz and Handelman245,246 point out that, as a result, the forces exerted by the zonule become more tangential to the lens surface and are hence less able to impart tension to the lens capsule.

A second geometrical possibility is that the distance between the attachment points of the anterior zonular fibres on the ciliary body and on the lens in the unaccommodated condition diminishes as a result of age-related growth in the lens or ciliary body. This would reduce the tension that the fibres are able to exert on the capsule and lens and hence the possible changes in lens power. Although the width of the circumlental space in the relaxed eye diminishes with age (and shows nasal/temporal asymmetry),228 the effect of this on the tension of the zonule may be compensated for by the axial movement of the attachment points of the zonular fibres to the lens capsule232,233,238 (Figure 11).

Figure 11.

Regression line fits for the changes with age in the width of the circumlental space in the relaxed eye (after Strenk, Strenk and Semmlow228), the distance from the lens equator of the attachment points of the zonular fibres (after Sakabe and colleagues233) and their sum


Accommodation is reasonably effective by the age of about four months247–249 and remains more than adequate for most purposes until the onset of presbyopia at the age of about 40 years. Although most current models of accommodation still draw their inspiration from the work of Helmholtz and his broad-brush view of the nature of the mechanism in terms of the interactions between the ciliary muscle, zonule and lens is generally accepted, the exact roles played by several of the ocular structures involved continue to be a matter of debate. In particular, the challenge of obtaining a full understanding of the optical and viscoelastic properties of the heterogeneous crystalline lens remains unmet. While many details remain controversial, the bulk of the available evidence supports the traditional view that lenticular growth and the associated changes in its viscoelastic properties dominate the development of presbyopia, although some contributions are probably made by changes in several other elements of the accommodative system. After the onset of presbyopia, there is evidence of a continuing decline in the performance of these non-lenticular elements, particularly the capsule. These further changes may reduce the effectiveness of surgical procedures designed to restore accommodation by simply compensating for lenticular changes, such as implantation of accommodating intraocular lenses or refilling of the lens capsule.4