Most organisations upgrading to LED lighting focus on energy savings, lux levels, and colour temperature. Very few ask about flicker. That is a significant oversight, because flicker in LED lighting is one of the most underreported causes of headaches, eye strain, fatigue, and reduced concentration in the workplace, and in the most serious cases, it can trigger seizures in people with photosensitive epilepsy.
This post explains what flicker is, why LED lighting is vulnerable to it, how to measure it, who is most at risk, and what responsible specification looks like. It is the first in our series on Lighting, Health and the Human Mind, a series examining the evidence base behind lighting's effect on human wellbeing in commercial environments.
Flicker is the rapid, repeated variation in the output of a light source. All light sources powered by alternating current have some degree of modulation in their output, the question is whether that modulation is fast enough, small enough, and regular enough to be imperceptible and harmless.
With older fluorescent technology, visible flicker was a well-known problem, particularly as tubes aged. The widespread shift to LED was partly driven by the expectation that flicker would be eliminated. In practice, it has not been. The quality of the LED driver, the component that converts mains electricity into the direct current the LED needs, determines whether a fitting flickers significantly or not. Low-quality or poorly specified drivers can produce substantial flicker even in otherwise well-performing luminaires.
Flicker is also not always visible to the naked eye. Visible flicker typically occurs at frequencies below around 60 to 100 Hz. But flicker above that threshold, flicker you cannot consciously detect, can still affect the visual system and has been linked to neurological symptoms in sensitive individuals. This is sometimes called invisible flicker or temporal light artefacts (TLAs), and it is the primary concern in modern commercial LED installations.
Two main metrics are used to characterise flicker in lighting. Understanding them matters when evaluating product data sheets or comparing specifications.
Percent flicker describes the depth of modulation in a light source's output, how much the light varies between its maximum and minimum levels. A light source with zero percent flicker has a completely stable output. A source with 100 percent flicker is switching fully on and off with each cycle.
The formula is: (Max − Min) ÷ (Max + Min) × 100.
IEEE 1789, the standard published in 2015 by the Institute of Electrical and Electronics Engineers titled Recommended Practices for Modulating Current in High-Brightness LEDs for Mitigating Health Risks to Viewers, sets out risk thresholds based on percent flicker at different frequencies. At 100 Hz, the typical frequency for a 50 Hz mains supply, IEEE 1789 recommends that percent flicker should not exceed 8% to remain in the low-risk zone. Above this threshold, the standard identifies increasing risk of biological effects.
SVM is a more recent metric, defined in IEC TR 61547-1 (most recently revised in 2020), which assesses the likelihood that a light source will create the stroboscopic effect, the phenomenon where moving objects under flickering light appear to strobe or freeze. This is particularly relevant in workplaces where machinery is in use, as it creates a genuine safety hazard.
An SVM value of 1.0 represents the threshold of visibility for an average observer. Values below 0.4 are generally considered acceptable for office environments. The EU's Ecodesign Regulation (EU) 2019/2020, which came into force from September 2021 for light sources, sets maximum SVM values for products placed on the European market, making this a regulatory requirement as well as a best-practice metric.
A third metric, PstLM (short-term flicker severity for lighting), applies to flicker caused by supply voltage fluctuations rather than the luminaire itself. It is less relevant to product specification but forms part of the IEC TR 61547-1 framework.
Flicker does not affect everyone equally. The evidence points to several groups who are significantly more vulnerable, and whose presence in a workforce creates a duty of care obligation for employers.
Photosensitive epilepsy (PSE) affects approximately 1 in 4,000 people in the general population, around 3% of all people with epilepsy. In people with PSE, exposure to flickering or flashing light within a critical frequency range can directly trigger a seizure.
The Epilepsy Society UK identifies the highest-risk frequency range as between 3 Hz and 60 Hz, with flashing between 16 and 25 times per second being the most likely to trigger a seizure. Critically, most LED mains-frequency flicker occurs at 100 Hz (double the 50 Hz supply frequency), which sits above the most dangerous range but is not automatically safe for all individuals. The Epilepsy Society notes that some people are sensitive to frequencies up to 60 Hz, and the broader range of 3–60 Hz should be treated as a zone of risk.
The research foundation here is substantial. Professor Arnold Wilkins of the University of Essex, one of the world's leading authorities on visual stress and flicker, reviewed the evidence on photosensitive epilepsy and light in his 1995 book Visual Stress (Oxford University Press), establishing flicker-induced seizure as a well-evidenced clinical phenomenon. His subsequent work, including a 2010 paper co-authored with colleagues and cited in IEEE 1789, directly addressed the risk profile of LED lighting and the need for standards-based controls.
Migraine affects approximately 10 million people in the UK, according to the NHS, making it one of the most prevalent neurological conditions in any workforce. People with migraine are frequently photophobic, abnormally sensitive to light, both during and between attacks.
Wilkins' work on visual stress, particularly his 2007 paper Prevention of Visual Stress and Migraine with Precision Spectral Filters, identifies flickering light as a significant trigger for both headache and the visual disturbances that precede migraine. His 2021 paper in the journal Neuro-Ophthalmology specifically notes that flicker "interferes with eye movements across text" and "causes headaches", with particular relevance to screen-based work environments.
A 2012 risk assessment paper published in the British Medical Journal's Injury Prevention supplement (Rider et al., 2012) catalogued the adverse effects of LED flicker as including "seizure, stroboscopic effects, migraine, exacerbation of repetitive behaviour in persons with autism, and general malaise", a useful summary of the breadth of the risk population.
Visual stress, also called Meares-Irlen syndrome or scotopic sensitivity, is a condition characterised by visual distortion, discomfort, and fatigue when reading or processing high-contrast visual environments. It is associated with, but distinct from, dyslexia, and is thought to affect between 5% and 20% of the population to varying degrees.
Wilkins' research demonstrates that this population is particularly sensitive to flicker, with flickering light exacerbating visual distortion and reducing reading speed and accuracy. In an office environment where reading, screen work, and document processing are central tasks, this is directly relevant to productivity and wellbeing.
Even for people without a specific sensitivity condition, sub-threshold flicker has measurable effects. The US Department of Energy's 2015 summary of the IEEE 1789 framework references evidence linking temporal light modulation to "neurological problems including epileptic seizure, headaches, fatigue, blurred vision, eyestrain, migraines, and reduced visual task performance" across the general population.
HSE guidance document HSG38 Lighting at Work specifically identifies flicker and stroboscopic effects as factors that employers must address as part of their lighting risk assessment obligations under the Workplace (Health, Safety and Welfare) Regulations 1992 and the Management of Health and Safety at Work Regulations 1999.
The irony of the LED transition is that it introduced a new and less visible category of flicker risk. Traditional incandescent and halogen lamps have a degree of thermal inertia, the filament does not cool and reheat instantly, which smooths out mains-frequency modulation naturally. Fluorescent lighting, when operating correctly with an electronic high-frequency ballast, also operates at very high frequencies, reducing the risk of flicker.
LEDs have no such inertia. They respond to changes in driving current almost instantaneously. This means the LED driver's quality is the sole determinant of the luminaire's flicker performance. A well-designed driver with proper filtering will produce a stable output. A poorly designed or budget-specified driver may produce significant modulation at 100 Hz, 200 Hz, or at other frequencies depending on the driver topology.
The market contains a very wide range of driver quality. Premium architectural LED products from reputable manufacturers typically specify flicker performance and can provide IEEE 1789 or IEC TR 61547-1 compliant data. Mid-market and contractor-grade products frequently do not, meaning flicker performance is simply unknown unless independently tested.
Dimming compounds the problem. Many LED dimming systems, particularly phase-cut (TRIAC) dimmers not designed for LED products introduce significant flicker at lower dim levels. A fitting that is flicker-compliant at full output may produce unacceptable flicker at 30% or 50% brightness. This is especially relevant in offices and hospitality spaces where dimming is used routinely.
Addressing flicker risk in a lighting specification requires explicit requirements, not assumptions. The following are the minimum expectations for any commercial project in which occupant well-being is a priority.
Under the Workplace (Health, Safety and Welfare) Regulations 1992, employers have a duty to ensure that lighting is suitable and sufficient for the workplace and does not pose a risk to health. Under the Management of Health and Safety at Work Regulations 1999, that duty extends to conducting risk assessments and taking preventive action.
The Equality Act 2010 adds a further layer of obligation where employees have a known condition, such as photosensitive epilepsy or migraine, that is materially affected by their working environment. In those cases, reasonable adjustments may extend to the specification and installation of low-flicker lighting in the affected employee's workspace.
HSG38 explicitly identifies flicker and stroboscopic effects as risks to be addressed in lighting risk assessments. Organisations that have not considered flicker as part of their workplace lighting review are, in the strictest reading of that guidance, operating an incomplete risk assessment process.
Flicker is not a fringe concern. It is a well-evidenced risk with direct implications for employee health, productivity, and employer liability. The science is clear, the standards exist, and the measurement tools are available. What has been missing is the habit of asking for the data and writing it into specifications.
The shift to LED has brought enormous benefits in energy efficiency and controllability. But it has also introduced a new class of flicker risk that was not present in well-maintained fluorescent installations. Specifying for flicker performance is not technically difficult or expensive when it is addressed at the design stage. It becomes expensive and sometimes impossible to resolve once discovered after installation.
In the next post in this series, we look specifically at Lighting and Epilepsy: What Employers and Designers Need to Know, examining the clinical picture of photosensitive epilepsy in more detail and what the Harding guidelines mean for specification decisions.
For related reading on glare and visual comfort in the workplace, see our guide to UGR and Glare Reduction in Offices. For controls and dimming considerations, see DALI vs 0–10V vs Smart Lighting.