A cataract is a clouded lens. The lens sits behind the pupil and focuses light onto the retina; when its proteins aggregate and scatter light instead of transmitting it, vision goes soft and dim, colours fade, and headlights and low sun produce disabling glare. It happens to nearly everyone eventually.
Cataract is the leading cause of blindness worldwide — 33 percent of global blindness in 2010, down from 39 percent in 1990 [1], and of the 216.6 million people with moderate or severe vision impairment in 2015, 52.6 million from cataract [2] [3] [4].
And this review is the hopeful one in the vision set. Glaucoma takes peripheral vision irreversibly. Macular degeneration takes central vision and, in its dry form, has no treatment that restores it. Cataract takes everything, gradually — and then a twenty-minute operation gives it back. The lens is the one part of the eye that can simply be replaced.
Which means the global burden is not a statement about biology. It is a statement about access. Cataract and uncorrected refractive error together caused 55 percent of blindness and 77 percent of vision impairment in adults over 50 in 2015 [2] — and both are correctable. More than half of the world's blindness is, in principle, a supply problem.
Start here: why lenses cloud
The lens is unusual tissue. It has no blood supply and no nerves, it is built from cells that lose their organelles to become transparent, and the proteins in those cells — crystallins — are never replaced. The ones you are born with must stay soluble and correctly folded for eighty years [5]. Crystallins double as molecular chaperones, and their chaperone function is part of what keeps the lens clear [6].
Age-related nuclear cataract is, in the words of one review's title, a problem where "oxidation is the key" [7] — the lens nucleus loses its protection against oxidative damage, protein thiols oxidise, crystallins cross-link and aggregate, and aggregates scatter light [8] [9]. The same ageing process stiffens the lens and causes presbyopia, the loss of near focus that arrives in the forties [10].
Three cataract types are distinguished because they behave differently and have different causes: nuclear (central, from the ageing changes above, producing gradual blur and sometimes a temporary myopic shift — the "second sight" that briefly lets someone read without glasses), cortical (spoke-like opacities from the periphery inwards), and posterior subcapsular (a plaque at the back of the lens, which disproportionately ruins reading and driving into sun because it sits at the focal point and worsens with pupil constriction).
The risk factors map onto those types. Type 2 diabetes raises overall cataract risk — a meta-analysis of 8 studies and 20,837 subjects found an odds ratio of 1.97 (95% CI 1.45–2.67) — and the association was significant for cortical (1.68) and posterior subcapsular (1.55) cataract but not for nuclear sclerosis (1.36, 95% CI 0.97–1.90, P = 0.070) [11] [12] [13]. Corticosteroids classically cause posterior subcapsular cataract, and even inhaled steroids have been associated with the risk [14]. Ultraviolet exposure and smoking are the other established contributors.
Pillar 1: measurement and diagnosis
The examination
Cataract is diagnosed by looking. Visual acuity is measured, then the lens is examined at the slit lamp, which projects a thin sheet of light through the eye so the opacity can be seen in cross-section and graded.
The grading standard is the Lens Opacities Classification System (LOCS III), which scores nuclear opalescence, nuclear colour, cortical and posterior subcapsular opacity against reference photographs [15]. Because it is a subjective comparison, considerable work has gone into objective alternatives — Scheimpflug densitometry correlates with LOCS grading [16] [17] [18], as does swept-source OCT [19] — and LOCS grade has been related to visual function measures [20].
The measurement that actually decides surgery
Grading the lens is not the same as deciding to operate. What matters is how much the cataract is interfering with the patient's life, and that is not read off a slit lamp. A dense cataract in someone who reads little may matter less than a modest posterior subcapsular opacity in someone who drives at night — and cataract's effect on driving has been studied directly, with crash risk and driving habits examined in older drivers [21]. Visual impairment is also associated with falls in older adults [22].
Ruling out the coexisting disease
This is the step that determines whether surgery will help, and it recurs throughout this review. Cataract is a disease of older eyes, and so are glaucoma and macular degeneration. An eye can have all three. Removing the cataract from an eye whose macula is already destroyed will brighten the image without restoring the sight — so the retina and optic nerve must be assessed before surgery is promised, and where the cataract is too dense to see through, that assessment has to be indirect. In the UK national database, 36.9 percent of cataract surgery cases had ocular co-pathology [23].
Before operating, one thing that is not needed: routine preoperative medical testing, which was shown not to add value [24]. Biometry — measuring the eye to choose the lens power — is needed, and its accuracy is the main determinant of the refractive result [25] [26] [27] [28].
Centerpiece: what the operation delivers, and what limits it
The age gradient, and which burden climbs fastest
The Rotterdam study screened 6,775 people aged 55 and over. Blindness prevalence rose from 0.1 percent at ages 55–64 to 3.9 percent at 85 and over; visual impairment from 0.1 percent to 11.8 percent [29].
The check the arithmetic was never given. The paper reports those four prevalences and never divides them. Across the same age span, blindness rises 39-fold and visual impairment 118-fold — impairment escalates three times faster than blindness. And the paper attributes the two rises to different diseases: above 75, macular degeneration was the major cause of the increase in blindness, whereas age-related cataract predominantly caused the increase in visual impairment. So in a population with good surgical access, the fastest-climbing burden is the reversible one — held at impairment rather than allowed to reach blindness. The same paper states the lever directly: adequate implementation of cataract surgery could reduce visual impairment by one third [29].
Globally the ordering reverses. In 2010 cataract caused 33 percent of blindness but only 18 percent of moderate or severe impairment [1] — it lands on the severe end. Same disease, opposite position in the severity distribution. What differs between the Dutch cohort and the world is whether an operation arrives before sight is lost.
The outcome, and its ceiling
A 1994 meta-analysis of 90 studies pooled 17,390 eyes for acuity and 68,316 for complications: 95.5 percent of eyes without pre-existing ocular comorbidity reached postoperative acuity of 20/40 or better (95% CI 95.1–95.9), and 89.7 percent of all eyes [30].
A 2015 UK national database collected 180,114 eyes from 127,685 patients prospectively across 28 sites: 94.6 percent without ocular co-pathology reached 0.30 logMAR or better — the same acuity — against 79.9 percent of eyes with co-pathology [23].
A second check, across two datasets and twenty-one years. The UK database also reports that 36.9 percent of its cases had co-pathology. Mixing its two stratum-specific results at its own co-pathology share gives 0.369 × 79.9 + 0.631 × 94.6 = 89.2 percent — against the 1994 meta-analysis's all-eyes figure of 89.7 percent. Half a percentage point apart, from independent datasets separated by two decades and a change of surgical era. Neither paper performs that mixture.
What that reconstruction means is the important part. Success in an otherwise healthy eye was already 95.5 percent in 1994 and is 94.6 percent now. Modern phacoemulsification has not raised it, because it was already at the ceiling — and the ceiling is set by other eye disease, not by the operation. All of the difference between 94.6 and 79.9 percent is comorbidity. That is why the pre-operative assessment above matters, and it is why a patient with macular degeneration should be told what cataract surgery will and will not do for them.
Three honest limits. Only the two extreme age bands are plotted on the left because those are the values Rotterdam reports; intermediate bands are not given and are not invented. "20/40 or better" is a functional threshold, not full restoration. And Rotterdam is a Dutch population with good surgical access — the age gradient in a setting without it looks different, which is the whole point of the global comparison.
Pillar 2: treatment
There is no drug
Nothing taken by mouth or dropped in the eye clears an established cataract. Antioxidant approaches follow from the oxidation mechanism [8] and have not produced a treatment. The definitive treatment is surgical, and it is one of the most frequently performed operations in medicine.
The operation
Phacoemulsification uses an ultrasonic probe through a small incision to break up and aspirate the cloudy lens, leaving the posterior capsule in place, into which an intraocular lens is implanted. It is typically done under local anaesthesia as a day case. Microincision variants have been compared against standard coaxial phaco [31], and outcomes remain good in the hands of residents in training [32].
The complications are uncommon and well quantified, which is what allows the operation to be offered so freely.
Posterior capsule rupture with or without vitreous loss occurred in 1.95 percent of the 180,114 UK cases — and was associated with a 42-fold higher risk of retinal detachment surgery within three months and an eight-fold higher risk of endophthalmitis [23]. That is the intraoperative event that matters, and its consequences are why it is tracked.
Endophthalmitis — infection inside the eye — is the feared complication, and its prevention is one of ophthalmology's clearest randomised results. The ESCRS study recruited 16,603 patients in a 2×2 factorial design of intracameral cefuroxime and topical levofloxacin. Twenty-nine patients developed endophthalmitis, 20 with proven infection, and absence of intracameral cefuroxime was associated with a 4.92-fold increase in risk [33] [34] [35]. A cheap antibiotic injected into the eye at the end of the case cuts the rate roughly five-fold.
Post-operative inflammation and macular oedema are managed with steroid and non-steroidal drops [36].
Choosing the lens
Monofocal lenses give one focal distance and excellent optical quality; the patient wears glasses for the other distance. Aspheric designs reduce spherical aberration [37].
Toric lenses correct astigmatism at the same time [38], which matters more than it sounds: in a large series of cataract surgery candidates, 22.2 percent had corneal astigmatism of 1.50 dioptres or higher [39].
Multifocal lenses aim to give both distance and near vision without glasses, and they carry a real trade-off. A Cochrane review compared multifocal against monofocal lenses after cataract extraction [40] [41] [42] — and two separate papers are devoted to dissatisfaction after multifocal implantation [43] [44], generally from glare, haloes and reduced contrast. Spectacle independence is bought with optical quality.
Where the capsule does not support a standard lens, fixation techniques exist [45] [46].
Posterior capsule opacification
The commonest late complication is not a failure of the lens but regrowth: residual lens epithelial cells migrate across the capsule left behind, and it clouds again [47] [48] [49]. Incidence has been systematically reviewed [50], and it depends on lens material and design [51] — a sharp rectangular posterior optic edge creates a discontinuous bend that blocks cell migration and substantially reduces it [52] [53] [54] [55]. It is treated in minutes with a Nd:YAG laser capsulotomy, in clinic, without entering the eye.
That sequence — a common late complication designed out by changing the shape of an implant edge, and the residue treated by an outpatient laser — is a good illustration of why cataract care works as well as it does.
The access problem
This is where the burden actually sits. Cataract surgical coverage measures the proportion of people needing surgery who have received it; effective coverage adds the requirement that the outcome was good. An analysis of 148 Rapid Assessment of Avoidable Blindness surveys across 55 countries established baseline estimates of both, and of the gap between them — the quality gap — against a World Health Assembly target of a 30-percentage-point increase in effective coverage by 2030 [56] [57] [58].
And coverage is not distributed evenly. A systematic review of 23 population-based surveys in low- and middle-income countries found 21 showed higher coverage among men, with men 1.71 times more likely to have had cataract surgery than women (95% CI 1.48–1.97). The authors estimate that blindness and severe visual impairment from cataract could be reduced by around 11 percent by closing that gap alone [59].
High-volume surgical models in low-resource settings have been developed and their outcomes measured [60] [61], and national cataract surgical output has been related to socioeconomic indices [62].
Pillar 3: what is unresolved
Scaling coverage, and the quality gap inside it. The distinction between coverage and effective coverage exists because operations with poor outcomes do not count [56] [57]. Raising volume without raising quality moves one number and not the other.
Femtosecond laser-assisted surgery automates several steps. Its introduction [63] [64], safety and outcomes [65] [66], effect on effective phacoemulsification time [67] and on corneal endothelial cell loss [68] have all been studied — and so has its cost-effectiveness, including in a multicentre randomised trial against standard phacoemulsification [69] [70]. Given a procedure already at a 95 percent success ceiling, the question a new technology has to answer is what it improves.
Better lenses. Extended-depth-of-focus and other designs try to deliver a range of vision without the dissatisfaction multifocals can cause [43] [44] [41].
Pharmacological prevention. Delaying cataract rather than removing it would change the global arithmetic more than any surgical improvement, and the oxidation mechanism suggests where to look [7] [8]. Nothing has worked yet.
Refractive precision. Most remaining dissatisfaction in otherwise successful surgery is refractive, and formula accuracy is still being improved [27] [28] [71].
Dig deeper in lmmol
The vision set is now three, and reading them together gives the pattern. Glaucoma takes peripheral vision silently and irreversibly; macular degeneration takes central vision, reversibly in its wet form and not at all in its dry one; cataract takes everything and gives it all back. All three coexist in older eyes, which is exactly the comorbidity that caps cataract surgery's success at 79.9 rather than 94.6 percent [23] — the reason the other two reviews matter to this one is clinical, not thematic. Cataract also competes with them in the global cause tables [1] [2] [3]. Type 2 diabetes and glycemic control is the systemic link: diabetes roughly doubles cataract risk, and specifically the cortical and posterior subcapsular types [11] [13] [12] — and a diabetic eye may need both cataract surgery and retinal treatment [72]. Osteoporosis connects through falls, since visual impairment in older adults is a fall risk factor [22] and a hip fracture is often what a delayed cataract operation ultimately costs. The full collection is at health.