Category: 3. Business

Wireless ISPs face a growing threat from LEO satellite providers like Starlink that can reach rural users with faster download speeds.

There are around 2,000 U.S. wireless internet service providers (WISPs) and about nine million Americans get their internet service from these companies, according to the Wireless ISP Association (WISPA).  Many of these WISPs are very small and provide service to just a few hundred customers. 

WISPs have become more prevalent over the past few years largely due to the introduction of vendor equipment that makes it possible to more cost-effectively deliver better coverage using unlicensed spectrum and commercial off-the-shelf hardware.

WISPs deliver their services using fixed wireless access (FWA) but they tend to be smaller and focused on certain markets such as rural areas or apartment complexes than the large telcos like Verizon, T-Mobile or AT&T, which also use FWA technology to deliver broadband services across the country. However, unlike the WISPs, these operators don’t consider broadband to be their primary business. 

Using Ookla’s Speedtest Intelligence® data, we examined the performance of eight of the larger U.S. WISPs—Etheric Networks, GeoLinks, NextLink Internet, Resound Networks, Rise Broadband, Starry, Unwired Broadband, and Wisper Internet — from Q1 2021 through Q2 2025. For those providers that offer both FWA and fiber, we categorized users with upload speeds under 100 Mbps as FWA customers to distinguish them from fiber users. While all eight of the WISPs that we monitored improved their median download speeds during that time period, their performance varies greatly. 

Key Takeaways

• Starry, which is being acquired by Verizon, delivered the highest median download speeds (202.25 Mbps in Q2 2025) of all eight U.S. WISPs that we studied. 
• GeoLinks uses a combination of licensed and unlicensed high-band spectrum as well as some mid-band spectrum, and it delivered the slowest median download speeds (22.74 Mbps in Q2 2025) of the WISPs we reviewed. Its users in the 75th percentile (those in the upper end of the typical speed range) experienced download speeds of 56.58 Mbps in Q2 2025. 
• Because of Starry’s faster speeds, the WISP was able to deliver the FCC’s minimum requirement for broadband speeds of 100/20 Mbps to 66.88% of Speedtest users in Q2 2025. 
• WISPs face a growing threat from low-Earth orbit (LEO) satellite providers like Starlink, which can reach rural users with download speeds that are often faster than WISPs. 
• To continue to compete  in the broadband space, WISPs need to find ways to secure more spectrum to avoid network congestion and interference.

The Many Flavors of WISPs

The performance of WISPs in the U.S. is under scrutiny right now because of recent changes that the National Telecommunications and Information Administration (NTIA) made to the Broadband Equity and Deployment (BEAD)  program. In June 2025 the NTIA revamped BEAD to provide a technology-neutral approach and prioritize cost-per-location.This means that instead of favoring fiber, other technologies such as low-Earth orbit (LEO) satellite and FWA can compete with fiber for BEAD funding. The revisions also include a rule to ensure that bids go to the lowest-cost bidders.

States revised their BEAD applications and re-submitted them using the new guidance. Early indications are that many states plan to use FWA for at least a portion of their BEAD eligible locations. Connected Nation, a non-profit that monitors the digital divide, found that states have awarded 11.7% of eligible locations to FWA providers, and many of those FWA providers are categorized as wireless ISPs (WISPs). 

We analyzed the performance of eight of the largest U.S. WISPs over several quarters from Q1 2021 until Q2 2025. However, it’s important to note that all of these companies vary greatly in terms of their spectrum holdings, their business models, their coverage areas, and their vendor equipment, which drives a large variance in performance outcomes. 

Nevertheless, it’s notable that all eight of the WISPs we monitored improved their median download speeds during that time period. They also improved their median upload speeds, but to a much lesser extent. 

Starry outpaced all the others and recorded the highest median download speeds. In Q2 2025 Starry’s median download speed was 202.25 Mbps, which is more than double that of the Resound Networks with a median download speed of 99.41 Mbps in Q2 2025. Starry also was nearly nine times higher in median download speeds than the slowest of the eight WISPs, GeoLinks, which had a median download speed of just 22.74 Mbps in Q2 2025. 

The eight WISPs and their coverage areas

Name	States where WISP operates	Spectrum used
Etheric Networks	California	2.4 MHz, 5.8 GHz unlicensed and 28 GHz licensed
GeoLinks	California, Arizona, and Nevada	unlicensed 69 GHz and 71 GHz, unlicensed 81 GHz
NextLink	Texas, Oklahoma, Illinois, Iowa, Kansas, and Nebraska	2.4 MHz, 5 GHz, and 6 GHz
Resound Networks	Texas, New Mexico, Arizona, Colorado, Oklahoma, Arkansas, Kansas	6 GHz unlicensed, 5 GHz unlicensed, and 3.65 GHz licensed
Rise Broadband	16 states including Colorado, Nebraska, Illinois, Iowa, Texas and Southern Wisconsin	unlicensed 5 GHz, unlicensed 3.65 GHz, licensed 2.5 GHz, and some TV white space spectrum at 470-698 MHz
Starry Broadband	Major cities such as Boston, Denver, Los Angeles, New York City and Washington, DC	37 GHz licensed, 24 GHz licensed, some 5 GHz unlicensed
Unwired Broadband	California	unlicensed 6 GHz
Wisper Wireless	Oklahoma, Kansas, Indiana, and Illinois	3.5 GHz (CBRS), 5.1 GHz, and maybe 6 GHz

Most WISPs struggle to deliver the FCC’s minimum broadband speeds to their customers 

All of the eight WISPs use a different configuration of spectrum licenses. Most are reliant upon some combination of low-, mid-, or high-band licensed and unlicensed spectrum. In addition, many have deployed fiber either as an alternative to their FWA service or to use to carry backhaul or middle-mile traffic. 

While using unlicensed spectrum means that a WISP can launch services quickly without having to purchase costly spectrum licenses, it also means that congestion and interference can result in the WISP having to carefully manage demand for their services. 

Using Speedtest data collected in Q2 2025 we compared the median download and upload speeds of the eight WISPs to determine what percentage of their Speedtest users were receiving the FCC’s minimum standard for fixed broadband speeds (100 Mbps downstream/20 Mbps upstream).   

Starry, which has mmWave spectrum licenses and uses proprietary equipment, is able to provide the FCC’s minimum standard for broadband to the highest percentage of users at 66.9%.  In contrast Rise Broadband, which primarily operates with unlicensed spectrum in the 5 GHz band and in the 3.55 GHz to 3.7 GHz bands (CBRS), but also uses some licensed spectrum in the 2.5 GHz band, is able to provide the FCC’s minimum requirement for broadband to just 6.7% of its users. 

WISPs	% of Speedtest users achieving wireless broadband speeds of 100/20 Mbps
Starry	66.9%
Resound Networks	41.5%
Wisper Internet	26.0%
NextLink	24.4%
Unwired	21.8%
GeoLinks	8.7%
Etheric	8.4%
Rise Broadband	6.7%

mmWave’s bigger pipe doesn’t always equal faster speeds

Starry, GeoLinks and Etheric all use some combination of high-band spectrum to deliver their FWA services. The benefits of this spectrum is it can deliver faster speeds and carry bandwidth-intensive applications. But it also requires line-of-sight or near-line-of-sight to work because of potential interference from buildings, trees, and even rain. 

Among the three providers that use mmWave spectrum we saw dramatic differences with Starry significantly outperforming GeoLinks and Etheric, which suggest that Starry has a greater penetration of mmWave spectrum among its customer base that is benefitting the WISP. 

Starry

Starry uses a proprietary technology with base stations that cover a radius of about one mile and its system operates on shared spectrum licenses in the 37.1, 37.3 and 37.5 GHz mmWave bands. It also acquired 104 licenses in the 24 GHz band that cover 51 partial economic areas. 

The company targets large apartment buildings with its service. Its setup consists of a rooftop base station that broadcasts a signal to multiple building-mounted receivers, allowing a single base station to serve dozens of buildings. Although it uses proprietary equipment it’s based upon modified 802.11ac/ax standards that takes advantage of the Wi-Fi chipset ecosystem.

The company, which is currently being acquired by Verizon, offers service to about 100,000 subscribers in apartment buildings in five markets; Boston, Denver, Los Angeles, New York/New Jersey, and Washington, D.C./Virginia.

Starry offers a variety of rate plans: $30 per month for up to 200 Mbps; $55 per month for up to 500 Mbps; and $75 per month for up to 1 Gbps. 

Ookla’s Speedtest® data shows that Starry has nearly doubled its median download speeds in its markets from 102.74 Mbps in Q1 2022 to 202.25 Mbps in Q2 2025. The company’s upload speed also increased, but not as dramatically from 52.29 Mbps in Q1 2022 to 54.34 Mbps in Q2 2025.  The company saw the biggest increase in speeds from Q1 2024 to Q2 2025, which is likely due to some network upgrades, including the deployment of the 2.0 version of its Comet receiver.  Starry said the upgrades would expand its coverage range as well as provide better spectral efficiency.  

GeoLinks 

GeoLinks uses local multipoint distribution services (LMDS) spectrum that it acquired from Verizon in 2021. Those 208 licenses are in the 29/31 GHz bands and cover several markets. However, GeoLinks currently offers service primarily in California and has a few deployments in Arizona and Nevada, but our Speedtest data samples were all collected from the company’s California deployment. 

The company recently used Intracom Telecom’s point-to-multipoint equipment to demonstrate multi-gigabit FWA using its 29/31GHz mmWave spectrum. In addition, it has indicated that it is interested in leasing its spectrum to other enterprises and operators that can then use its spectrum holdings to develop their own FWA services. 

GeoLinks offers a variety of price plans: $25.99 per month for speeds of 10/10 Mbps; $38.99 per month for 25/10 Mbps; $45.99 per month for 30/30 Mbps; and $69.99 per month for speeds of 100/25 Mbps. The company’s web site indicates that the $45.99 per month plan that delivers 30/30 Mbps is the most popular plan with its customers. 

Speedtest data shows Geolinks delivering median download speeds of just 22.74 Mbps in Q2 2025 with 75th percentile download speeds of 56.58 Mbps. Its users experience median upload speeds of 19.82 Mbps in Q2 2025.  

Etheric Networks

Etheric Networks provides FWA service to the California Bay Area. The company has a fiber ring stretching from San Francisco to Monterey, California that connects its FWA towers and eight data centers. Etheric uses a mix of spectrum including unlicensed 2.4 GHz and 5.8 GHz spectrum. However, in 2024 Etheric partnered with BroadbandOne to leverage BroadbandOne’s 28 GHz mmWave spectrum. The company said this partnership will allow it to enhance its connectivity and serve more rural and agricultural areas. 

The company offers three residential price plans: $79 per month for speeds up to 100 Mbps; $99 per month for speeds up to 250 Mbps and $169 per month for 1 Gbps speeds. 

Speedtest data shows Etheric has nearly doubled its median download speeds from 21.34 Mbps in Q1 2021 to 41.09 Mbps in Q2 2025. Its users in the 75th percentile (those in the upper end of the typical speed range) saw speeds of 65.45 Mbps in Q2 2025.The company’s median upload speeds also increased over time from 13.6 Mbps in Q1 2021 to 29.5 Mbps in Q2 2025. 

WISPs make the most of mid-band with CBRS licenses

Many WISPs take advantage of the mid-band CBRS spectrum, which is a 150 MHz shared spectrum in the 3.5 GHz to 3.7 GHz band that allows for flexible use by three different groups that are managed by a Spectrum Access System (SAS). The SAS can dynamically grant access to different users. The band is shared by these three parties: incumbent users such as the U.S. Navy that have priority access to the band; licensed users with Priority Access Licenses (PAL) that have exclusive use of a portion of the band in a specific geographic location; and the General Authorized Access (GAA) group who can access the spectrum but have no protection from interference from the other two groups.  

Several of the WISPs we analyzed deploy their services in the CBRS spectrum and primarily use the GAA portion of the band. Others have acquired CBRS PAL and some use a combination of both. Some WISPS also use unlicensed bands such as 5 GHz. 

NextLink

Nextlink spent $28.4 million in FCC’s Auction 105 to purchase over 1,100 CBRS PAL licenses covering 491 counties in eleven states including Texas, Oklahoma, Kansas, Nebraska, Iowa, Minnesota, Wisconsin, Indiana, Wyoming, and Missouri. The company uses that spectrum to deliver its FWA service to its more than 100,000 subscribers (as of August 2025).  NextLink also has deployed fiber to more than 100,000 locations and has 20,000 fiber customers. 

Nextlink secured Connect America Fund II funding and participated in the FCC’s Rural Digital Opportunity Fund so much of its FWA expansion has been driven by those commitments. In August Nextlink said it has completed five of the six states as part of its CAF II funding and is halfway through its RDOF buildout. 

The company offers a variety of FWA plans: The Next50, which offers up to 50 Mbps speeds for $30 per month; the Next100 that offers speeds up to 100 Mbps for $40 per month; The Next300 that offers speeds up to 300 Mbps for $60 per month; and the Next500 that offers speeds up to 500 Mbps for $75 per month. 

Speedtest data shows NextLink has more than tripled its median download speeds from 19.45 Mbps in Q1 2021 to 68.47 Mbps in Q2 2025.  The WISP also increased its median upload speeds significantly from 4.72 Mbps in Q1 2021 to 18.26 Mbps in Q2 2025. NextLink users in the 75th percentile (those in the upper end of the typical speed range) get much higher speeds of 122.88 Mbps in Q2 2025. 

Resound Networks

Resound Networks provides FWA service in Texas, New Mexico, Arkansas, Arizona and Oklahoma and uses Tarana Wireless gear in the unlicensed 5 GHz and 6 GHz spectrum bands. It also offers fiber service in some locations and is planning to expand its fiber footprint. Like many WISPs, Resound is focused specifically on rural communities that have historically been overlooked by larger ISPs. In 2022 the company was awarded $303 million through the FCC’s RDOF program to deliver FWA and fiber to 214,000 rural locations. 

Resound offers both residential and enterprise rate plans. Its residential plans start at 75 Mbps for $55 per month and go up to 1 Gbps for $130 per month. 

The company’s customers experienced a steady increase in their download and upload speeds from mid-2023 until Q2 2025 from a median download speed of 38.94 Mbps in Q3 2023 to 99.41 Mbps in Q2 2025.  Its users in the 75th percentile (those in the upper end of the typical speed range) experienced an even greater climb in download speeds from 62.99 Mbps in Q3 2023 to 190.76 Mbps in Q2 2025.  During this time period Resound was expanding its network. 

Rise Broadband

Rise Broadband claims to be the country’s largest WISP with around 200,000 customers. It may also be one of the longest living WISPs because it dates back to 2006 when it started as JAB Broadband and its goal was to consolidate many of the country’s smaller WISPs to create one big WISP with a large footprint. 

Today Rise offers FWA service in16 states, mostly in the Midwest. Rise offers service primarily in rural areas and it uses a mix of unlicensed spectrum in the 5 GHz band and in the 3.55 GHz to 3.7 GHz bands (CBRS), but also uses some licensed spectrum in the 2.5 GHz band, to deliver its service. 

Like NextLink, the company is actively deploying fiber in addition to FWA. The company’s strategy is to deploy FWA initially to capture market share and then roll out fiber to the densest FWA coverage areas. 

Rise’s price plans start as low as $30 per month for 50 Mbps and reach up to 400 Mbps for $55 per month. 

Rise users logged median download speeds of 42.58 Mbps in Q2 2025, which is a significant jump from Q1 2021 when users experienced median download speeds of just 16.01 Mbps. Rise’s users  in the 75th percentile (those in the upper end of the typical speed range) were able to achieve download speeds of 65.97 Mbps in Q2 2025.  The company’s median upload speeds also increased from 4.05 Mbps in Q1 2021 to 18.38 Mbps Q2 2025. Rise saw a big jump in median upload speeds between Q2 2022 when users logged median upload speeds of 5.86 Mbps and Q3 2022 when users experienced median upload speeds of 13.68 Mbps. 

Wisper Internet

Wisper Internet offers FWA in six midwestern states including Illinois, Missouri, Kansas, Oklahoma, Arkansas and Indiana. The company uses unlicensed spectrum in the 5 GHz, and a mix of unlicensed and licensed spectrum in the 2.5 GHz and 3.65 GHz bands.  Like NextLink and Rise, the company also has deployed fiber in a few select areas. 

Wisper offers a variety of rate plans including 25 Mbps for $70 per month; 50 Mbps for $75 per month; 100 Mbps for $80 per month; 200 Mbps for $110 per month and 400 Mbps for $140 per month. 

Similar to the other WISPs, Wisper’s median download speeds increased over time but it increased dramatically from Q3 2023 to Q2 2025 when its median download speeds increased from 33.74 Mbps to 52.90 Mbps. Likewise, the download speeds for users in the 75th percentile also increased, climbing from 55.12 Mbps in Q3 2023 to 107.90 Mbps in Q2 2025. This jump in speeds was likely due to  Wisper’s deployment of additional FWA gear from Tarana Wireless on 180 more towers in its footprint. 

Unwired

Unwired Broadband provides FWA coverage in rural and underserved areas in central and northern California. The company said it has a network of more than 200 towers and a coverage area of about 17,000 square miles. Besides FWA, Unwired also provides some fiber service but It’s early in its deployment process. 

Unwired uses a combination of licensed and unlicensed spectrum to deliver its FWA service, including the licensed 2.5 GHz band and the unlicensed 6 GHz band. 

The company offers both business and residential FWA service and its pricing starts at $59.99 per month for 100 Mbps. 

Unwired users experienced increases in download and upload speeds over time but between Q3 2024 and Q4 2024 the jump was more dramatic. Median download speeds jumped from 27.22 Mbps in Q3 to 44.25 Mbps in Q4. Similarly median upload speeds increased from 9.7 Mbps in Q3 2024 to 15.9 Mbps in Q4. 

WISPs’ performance is improving but competitive threats lurk 

Although the WISPs we studied are improving their networks and delivering better performance for their customers, the broadband market is rapidly changing. In the past many WISPs, particularly those in rural areas, faced little or no competition. But that’s no longer the case. 

As LEO satellite constellations such as Starlink become more powerful and more prevalent (Amazon’s Kuiper now has 153 satellites in orbit and is expected to launch late this year), WISPs will face growing competition from these companies. 

A recent Ookla report on Starlink found that Starlink’s network saw its median download speeds nearly double from 53.95 Mbps in Q3 2022 to 104.71 Mbps in Q1 2025, making its median download speeds on par or better than seven of the eight WISPs we reviewed (Starry was the only exception). With Starlink residential price plans starting around $80 per month, the company’s introductory price plan is a bit more expensive than some introductory price plans from WISPs but Starlink is aggressively promoting its services and offering large discounts on its equipment to entice new customers. 

To continue to play in the broadband space, WISPs need to try to secure more spectrum–licensed or unlicensed— to avoid network congestion and interference and also  invest in network upgrades so their services remain competitive. 

 To find out more about Speedtest Intelligence® data and insights, visit our website.  

Study design and participants

This study utilized a quasi-experimental, pre-test/post-test design to compare the learning outcomes of nursing students exposed to Simulated Clinical Immersion (SCI) against those in a traditional, lecture-based control group. This design was selected as a methodologically sound approach to assess the comparative effectiveness of teaching strategies where randomized controlled trial (RCT) allocation was deemed impractical due to logistical constraints within the academic curriculum. The target population comprised undergraduate nursing students. The final sample size was N = 128.

Sample size and power calculation

A formal power analysis was conducted a priori using (Specify the Software Used, e.g., G*Power) to determine the requisite sample size. Based on a moderate expected effect size (e.g., Cohen’s d = 0.50) observed in similar simulation-based education studies (Cite relevant prior studies here), with a statistical power (1−β) set at 0.80 and a significance level (α) of 0.05, the calculated minimum sample size was (Specify the required number, e.g., 102). The enrolled sample of N = 128 was therefore considered highly appropriate and sufficient to detect statistically significant differences between the groups.

Educational content and allocation

All participants received the same standardized didactic content on palliative care and end-of-life (EOL) principles for older adults prior to the intervention phase. This curriculum included: pain and symptom management, ethical and legal considerations in EOL care, and therapeutic communication with older patients and their families. This ensured equivalence in theoretical foundational knowledge across both groups.

Participants were assigned to either the intervention group (Simulated Clinical Immersion – SCI) or the control group (Traditional Lecture) via alternate allocation (Non-random assignment based on class section or enrollment order), aligning with the study’s quasi-experimental classification.

Measurement tools

Four instruments were administered as both pre-tests and post-tests to assess the core learning outcomes: knowledge, self-efficacy, and preparedness for EOL care. All instruments demonstrated high internal consistency in the current study.

Demographic data sheet

A researcher-developed demographic questionnaire was used to collect participants’ background information, including age, gender, prior exposure to palliative care, and previous experience with simulation-based learning. Content validity was confirmed through expert review by faculty specializing in palliative care education to ensure the clarity and relevance of the items.

EOL knowledge

The scale was adapted from Gellis et al. [11], to specifically target clinical knowledge in geriatric palliative care. While the original tool evaluated interprofessional team communication, the modified version focused exclusively on assessing knowledge related to EOL symptom management and ethical principles in older adults. The adapted questionnaire will be included in an appendix for transparency.

Clinical self-efficacy scale

This scale was adapted from two separate established tools to provide a comprehensive measure of students’ confidence in providing EOL care. The internal consistency of this adapted version in the current sample was confirmed by a high Cronbach’s α 0.91, supporting its reliability for this study’s context.

Simulation Effectiveness Tool (SET); The SET, developed by Elfrink et al. [12], assessed students’ perceptions of the simulation’s educational effectiveness. The tool includes 13 Likert-scale items addressing three key domains: learning outcomes, confidence building, and clinical relevance. The SET is a widely recognized and validated instrument for evaluating simulation-based learning. In this study, construct validity was maintained by ensuring item alignment with the goals of the SCI intervention. The tool yielded a Cronbach’s α of 0.89 during pilot testing, reflecting high internal consistency.

The pilot study used for initial testing of the instruments included a sample size of (N = 25) students. The internal consistency values (α) for all four instruments in the final N = 128 sample were consistently α > 0.80 confirming their excellent psychometric properties.

Ethical considerations

This study received approval from the Ethical committee of faculty of nursing at Zagazig University (Approval No. 130–2024), ensuring ethical compliance and protecting all participants’ rights. This approval, a testament to our unwavering commitment to upholding ethical standards, was crucial in the research process. Informed Consent: Participants were fully informed about the study and gave written consent before joining. It’s important to note that participation was entirely voluntary, and students had the right to withdraw at any time without any consequences, ensuring their security and control.

Intervention group: simulated clinical immersion (SCI)

The SCI experience was structured across three core phases (pre-briefing, simulation, and debriefing).

Simulation Scenario and Procedures. Each simulation session involved 8 students. 4 students acted as active participants (rotating roles such as primary nurse, communication lead, and assistant nurse), while 4 students served as active observers using a structured checklist and contributing to the debriefing phase. The scenario focused on a critically ill older adult presenting with an acute exacerbation of (Specify the clinical condition, e.g., advanced Chronic Obstructive Pulmonary Disease) requiring immediate symptom management and a goals-of-care discussion with the family. The session structure adhered to best practices in simulation reporting guidelines (Referencing Table 1 for healthcare simulation research).

Debriefing process

The debriefing session (approximately 30–40 minutes) immediately followed the simulation. It was led by faculty experts with specialized knowledge in geriatric care, EOL principles, and simulation-based learning. The process employed a structured debriefing model (Specify the model used, e.g., Plus-Delta or GAS) to facilitate student reflection and guided learning. Feedback was standardized across all sessions using a faculty checklist to ensure consistency in content, duration, and focus on the targeted competencies (e.g., pain assessment, ethical reasoning, therapeutic communication) regardless of the faculty member leading the session. The faculty instructors had an average of 7 years of experience in clinical nursing and 3 years of experience in simulation pedagogy.

Control group activities

The control group received only didactic, lecture-based instruction covering the identical learning objectives and theoretical content as the SCI group. The sessions included instructor-led presentations, assigned readings, and large group discussions but excluded any form of clinical simulation or high-fidelity experiential learning

Data analysis

Data were analyzed using SPSS version 26. Descriptive statistics summarized participant demographics, while independent and paired t-tests, along with chi-square tests, evaluated differences within and between groups. Effect sizes (Cohen’s d) quantified the practical significance of findings, interpreted using standard thresholds. Statistical significance was set at p < 0.05. The analysis revealed large effect sizes, underscoring the substantial impact of the simulation intervention on knowledge, self-efficacy, and perceived simulation effectiveness.

The latest take-home grocery sales in Ireland rose by 5.5% in the four weeks to 2 November 2025, according to our latest data. Despite prices continuing to rise, shoppers spent an additional €64 million on groceries and were out in-store more often over the four weeks, boosting overall sales by €5.8 million. 

Grocery price inflation is now standing at 6.06%, falling slightly from last month by 0.5%. Irish shoppers, however, are in full flow for the festive season as they start to stock up on seasonal favorites.

It’s that time of year again as Christmas ads light up our screens and supermarkets fill their shelves with festive fare, signalling that the countdown to the big day has begun. Shoppers are already getting into the spirit, spending an additional €8 million on boxed chocolates compared with last month.

Retailers are acutely aware of the financial pressures that many households are under right now, especially with this year’s Budget approaching, and are keen to show how they’re delivering value for money with a big focus on promotions. Nearly a quarter (22%) of all grocery sales are on promotion, up nearly 5.9% versus this time last year. Promotional activity typically intensifies in the lead-up to Christmas, so this is likely to persist well into December.

Shoppers seeking value and quality

Own label saw strong growth over the last 12 weeks, up 6.3%, with shoppers spending an additional €99 million on these ranges versus last year. Premium own label ranges continue to see strong growth, up 15.3%, with shoppers spending an additional €18.5 million on these ranges compared to last year.

Brands continue to grow behind the total market with growth slowing to 5.2% in the last 12 weeks, but Irish shoppers still spent an additional €85.4 million on branded products. 

Brands currently hold 48.2% value share of the total market, the highest share since February, showing how shoppers are more likely to treat themselves in the lead-up to Christmas. Own label holds 46.2% value share.

Shoppers are usually looking for both value and quality – not just the lowest price – particularly at Christmas. I think we all like to indulge a little at this time of year and, despite the ongoing cost-of-living crisis, many have managed to find a balance by choosing retailers’ premium own-label ranges as an affordable way to treat themselves.

Category increases included frozen fruit, low alcoholic drinks, hot beverages, confectionary sweets, chocolate, sweet spreads and savoury snacks, all of which grew ahead of the total market in the latest 12-week period.

Retailer and channel performance

Online continued to grow, at a significantly faster rate compared to last month, up 6.4% year-on-year to take 5.6% value share of the market. Shoppers spent an additional €13.1 million online during the period, helped by an influx of new customers who contributed €5.8 million to overall performance – more than double last month. Over 18% of Irish households bought their groceries online during this time. 

Dunnes holds 24.6% market share, up on the last 12-week period, with sales growth of 6.1% year-on-year. Larger and more frequent trips contributed an additional €26.2 million  to their overall performance. 

Tesco holds 23.8% of the market, with value growth of 7.7% year-on-year. Shoppers increased their trips to stores by 0.8%  and, together with new shoppers, contributed an additional €29.8 million to the grocer’s overall performance.

SuperValu holds 19.4% of the market with growth of 4.1%. Consumers made the most shopping trips to this grocer, averaging 24.1 trips over the latest 12 weeks. SuperValu recruited new shoppers to store over the latest 12 weeks which contributed an additional €11.4 million to their overall performance.

Lidl holds 14% of the market with growth of 9.6%, the fastest growth among all retailers once again. Lidl also saw shoppers pick up more volume in store, up 2.1%,  contributing an additional €9.5 million to overall performance. 

Aldi holds 11.2% market share, up 3.8%. Increased store trips and new shoppers drove an additional €9.9 million in sales.

Key Messages

What is already known on this topic

Drug-induced pigmentary changes are increasingly recognized but remain poorly characterized in large-scale surveillance studies. Previous literature has focused primarily on case reports or small series, and labeling for many implicated drugs is incomplete.

What this study adds

This FAERS-based pharmacovigilance analysis identified strong disproportionality signals linking numerous widely used drugs to skin hyperpigmentation and hypopigmentation. Several high-signal agents, including ribociclib and amlodipine, lack pigmentary warnings in their product labeling.

How this study might affect research, practice or policy

The findings support the need for enhanced dermatologic monitoring for high-risk drugs with pigmentary safety signals. Drug-induced pigmentary changes may indicate therapeutic repurposing potential in pigmentary skin disorders, warranting further translational research.

Introduction

Adverse drug reactions (ADRs), though increasingly prevalent and clinically impactful, remain underrecognized when presenting as skin pigmentary disorders, despite their visible manifestations and psychosocial consequences.^1,2 Drug-induced pigmentary changes account for approximately 20% of all cases of acquired pigmentation.³ The clinical heterogeneity and overlap of these pigmentary changes with common dermatoses-such as post-inflammatory pigmentation, melasma, or vitiligo-pose diagnostic challenges, often leading to misdiagnosis or delayed recognition. Drug-related hyperpigmentation and hypopigmentation are best managed by discontinuation or substitution of the culprit medication.⁴ Additional measures include rigorous photoprotection and topical agents such as hydroquinone for hyperpigmentation, whereas hypopigmentation may respond to topical corticosteroids or calcineurin inhibitors, phototherapy, or surgical techniques (eg, grafting techniques). Timely recognition of these manifestations is critical for mitigating drug-induced pigmentary sequelae.⁵

Drugs can induce pigmentary disorders via distinct pathways, resulting in either hyperpigmentation or hypopigmentation.⁶ Common drugs that lead to pigmentary changes include alkylating/cytotoxic agents, analgesics, antiarrhythmics, and anticoagulants.³ Drug-induced cutaneous hyperpigmentation is typically non–immune-mediated and arises through mechanisms such as melanin accumulation, drug deposition, novel pigment formation, and iron deposition. Melanin increase may result from melanocyte stimulation, drug-melanin complex formation, or inflammation, while vascular injury may lead to iron or pigment granule deposition within dermal macrophages.³ In contrast, drug-induced hypopigmentation is more often associated with immune-mediated or cytotoxic mechanisms that impair melanocyte function or viability. The principal pathways include decreased melanocyte density, inhibition of melanogenesis, and disruption of melanosome transport. Topical or intralesional corticosteroids may also cause localized hypopigmentation by suppressing melanocyte activity, typically near the site of application or injection, and often with partial reversibility.⁷

This study leverages the FDA Adverse Event Reporting System (FAERS)-the largest global pharmacovigilance database-to systematically identify drugs associated with skin pigmentary changes.⁸ Through an in-depth analysis of drug-related adverse event (AE) data from the FAERS database, this study aims to enhance clinicians’ vigilance and diagnostic awareness regarding drug-induced pigmentary disorders, thereby facilitating timely recognition and intervention in routine dermatologic practice. Moreover, identifying specific agents associated with pigmentary changes may offer novel insights for drug repurposing and optimization of therapeutic strategies in clinical pharmacology.

Methods

Data Source

This study was conducted between April and June 2025, with three investigators involved in the analysis of the FAERS dataset. This pharmacovigilance study was conducted based on data extracted from the FAERS, a publicly available database designed to support post-marketing safety surveillance for drugs and biologics. The FAERS database contains spontaneous reports of AEs voluntarily submitted by healthcare professionals, pharmacists, consumers, and other reporters. These reports provide detailed information on patient demographics, suspected medications, therapeutic indications, AEs, clinical outcomes, and other relevant clinical data. We utilized OpenVigil 2.1, a validated pharmacovigilance analytics platform (http://h2876314.stratoserver.net:8080/OV2/search/), to access and analyze the FAERS dataset. OpenVigil 2.1 performs data import, cleaning, and statistical processing directly on FAERS data and provides disproportionality analysis outputs, including the Reporting Odds Ratio (ROR). The version of OpenVigil used in this study contains AE reports from the first quarter of 2010 to the fourth quarter of 2024.⁹

Drug Identification

AEs were classified according to the Medical Dictionary for Regulatory Activities (MedDRA, version 24.0), using the Preferred Term (PT) level. We retrieved reports associated with two PTs: “skin hyperpigmentation” and “skin hypopigmentation”. To identify the suspect drugs related to these pigmentary changes, we extracted all reports containing these two PT terms in the AE. For consistency in drug naming, we standardized all drug entries to their generic names. Brand names were mapped to generics using the DrugBank database (https://go.drugbank.com). Reports with unidentified or ambiguous drug names were excluded after manual review.

Statistical Analysis

We conducted descriptive statistical analyses to characterize reports related to pigmentary disorders, including reporting year, patient age group, sex, country of origin, and the top five associated indications. The primary analytical method employed was disproportionality analysis–a widely accepted pharmacovigilance approach for detecting potential drug safety signals. Common disproportionality metrics include the ROR, Proportional Reporting Ratio, Information Component, and Empirical Bayes Geometric Mean, each based on distinct statistical assumptions and detection thresholds. In this study, we applied the ROR, a well-established and interpretable metric. ROR compares the odds of reporting a specific AE with a drug of interest to the odds of reporting the same AE with all other drugs in the database. A drug–AE pair was considered to yield a statistically significant signal if the lower bound of the 95% confidence interval (CI) for the ROR exceeded 1 and the number of reports was ≥3. Data processing, summarization, and visualization were performed using Microsoft Excel 2019 and R software (version 4.5.0).

Results

Demographic Characteristics

A total of 9594 hyperpigmentation-related AE reports were identified (see Table 1), with the majority from the United States (62.6%). The most common indications were psoriasis (5.9%) and magnetic resonance imaging (4.4%). The age distribution showed that most cases were in individuals aged 41–64 years. Hyperpigmentation-related AE reports peaked in 2023, as shown in Figure 1. A total of 2,397 hypopigmentation-related AE reports were identified (Table 2). Most cases involved individuals aged 41–64 years and ≥65 years. The top indications were malignant melanoma (13.8%) and hypertension (12.9%). The majority of reports came from the United States (46.5%), followed by Germany (13.3%). Reports steadily increased from 64 in 2010 to a peak of 352 in 2023, with a slight decline to 249 in 2024 (Figure 1).

Table 1 Clinical Characteristics of Reported Skin Hyperpigmentation

Table 2 Clinical Characteristics of Reported Skin Hypopigmentation

Figure 1 Annual reporting trends of drug-induced skin pigmentation disorders in FAERS, 2010–2024. The Orange line represents annual reports of hyperpigmentation, while the blue line denotes hypopigmentation. Hyperpigmentation cases were consistently more frequent and exhibited substantial year-to-year variation. In contrast, hypopigmentation reports were fewer and showed a steadier trend over time.

Disproportionality Signals

A total of 5,329 hyperpigmentation signals were identified based on the ROR threshold. The top 50 drugs with the highest signal strength are summarized in Table 3. Gadolinium-based contrast agents, notably gadodiamide (ROR 83.42), antibiotics such as minocycline (ROR 115.66) and clofazimine (ROR 53.30), topical agents including bimatoprost (ROR 105.36) and adapalene (ROR 20.25), pigmentation axis modulators represented by setmelanotide (ROR 1506.82), antimetabolites like mechlorethamine (ROR 65.99), and targeted therapies such as ribociclib (ROR 60.20) demonstrated the strongest disproportionality signals. Notably, certain drugs, including dupilumab and capecitabine, lack labeling for hyperpigmentation risks in their package inserts.

Table 3 Top 50 Drugs for Signal Strength of Skin Hyperpigmentation

A total of 1,054 hypopigmentation signals were identified based on the criteria of ROR. The top 50 drugs with high signal strength are listed in Table 4. These drugs were categorized into four major pharmacological classes: anticancer and targeted therapies, such as ribociclib (ROR 60.20), which exhibited one of the highest signal strengths; topical corticosteroids and dermatologics, including triamcinolone (ROR 37.20); immunomodulators and biologics, represented by dupilumab (ROR 3.46) and secukinumab (ROR 3.47); miscellaneous systemic drugs, such as dimenhydrinate (ROR 44.55). Notably, the majority of these drugs do not include skin hypopigmentation as a documented ADR in their package inserts.

Table 4 Top 50 Drugs for Signal Strength of Skin Hypopigmentation

Among hyperpigmentation-related events, the most frequently reported agents were bimatoprost (n = 567) and dupilumab (n = 210) (Figure 2). For hypopigmentation, dupilumab (n = 84) and letrozole (n = 61) accounted for the highest number of reports (Figure 2). Notably, several agents appeared in both hyperpigmentation and hypopigmentation datasets, including dupilumab, secukinumab, triamcinolone, cholecalciferol, clobetasol, hydroxychloroquine, and prednisone.

Figure 2 Comparative frequency of drugs associated with skin hyperpigmentation (left) and hypopigmentation (right) in FAERS, 2010–2024. Orange and blue bars represent the number of reports for drug-induced hyperpigmentation and hypopigmentation, respectively. Drug names are ranked by report frequency within each category; numeric values indicate absolute counts. An asterisk (*) denotes drugs reported in both pigmentation directions.

Discussion

While prior research on drug-induced pigmentary disorders has primarily relied on systematic reviews or isolated case reports, our analysis fills a crucial gap by utilizing large-scale pharmacovigilance data. Our findings reveal that reports of drug-induced hyperpigmentation were significantly more frequent than those of hypopigmentation. This discrepancy may stem from intrinsic mechanistic differences—hyperpigmentation often arises through more direct and common biological pathways,³ whereas hypopigmentation tends to involve rarer and more complex mechanisms—as well as from external reporting biases, such as variations in clinical visibility and diagnostic attention.⁷ Furthermore, inappropriate or excessive use of certain drugs, particularly those associated with long-term or high-dose exposure, may further widen the disparity between hyperpigmentation and hypopigmentation reports. The observed temporal pattern differences may partly reflect baseline reporting effects, with hyperpigmentation showing greater variability and hypopigmentation appearing relatively stable due to its lower baseline reporting frequency.

Among hyperpigmentation cases, females were disproportionately represented compared to males, whereas hypopigmentation exhibited no significant gender difference. This observation aligns with previous literature indicating a female predominance in drug-induced hyperpigmentary disorders, potentially attributable to estrogen’s stimulatory effects on melanogenesis in human melanocytes and the synergistic interactions of elevated α-melanocyte-stimulating hormone levels with estrogen.¹⁰ Both hyperpigmentation and hypopigmentation were predominantly reported in middle-aged and elderly patients, likely due to pre-existing imbalances in melanin metabolism associated with chronological skin aging and cumulative photodamage. This age distribution is consistent with previous clinical studies, such as the series by Giménez García and Carrasco Molina, which reported a mean patient age of 63 years (SD = 16; range, 34–86 years), with half of the affected individuals being over 60 years old.¹¹

Among drugs associated with skin hyperpigmentation, minocycline, bimatoprost, and setmelanotide warrant particular attention based on their robust signal strength. Minocycline, a tetracycline derivative widely used for acne and other inflammatory dermatoses, showed a notably high signal for hyperpigmentation in our study (ROR = 115.66; n = 199). Previous studies have suggested that approximately 3–15% of patients receiving long-term, high-dose minocycline therapy (defined as >1.5 years and cumulative doses >70–100 g) may develop cutaneous hyperpigmentation. This occurs due to the dermal deposition of melanin–iron–minocycline complexes.¹² Four distinct clinical patterns have been reported, often mimicking sunspots or age-related hyperpigmentation, leading to underrecognition.¹³ Therefore, regular dermatologic monitoring is advisable in patients receiving minocycline treatment. Early recognition can help prevent progression, and clinicians should inform patients of this potential risk prior to long-term use.

Bimatoprost is a synthetic prostaglandin F2α analog commonly used in the treatment of glaucoma and eyelash hypotrichosis.¹⁴ In our analysis, bimatoprost was associated with a remarkably strong hyperpigmentation signal (ROR = 105.36; 95% CI: 96.01–115.62). Consistent with this finding, multiple clinical studies have reported periorbital and eyelid hyperpigmentation following long-term ocular administration of bimatoprost, typically emerging within 4–6 weeks of treatment.^14–16 Mechanistically, bimatoprost has been shown to stimulate melanogenesis by upregulating tyrosinase activity via activation of prostanoid FP receptors, leading to increased melanin synthesis and deposition, particularly in periorbital regions. This pigmentary effect is typically reversible upon drug discontinuation. Interestingly, this pharmacologic property has also prompted its off-label use in repigmentation therapy for localized vitiligo and hypopigmented scars, suggesting a potential role in pigmentary disorders beyond its approved indications.¹⁷

Setmelanotide exhibited the strongest signal for hyperpigmentation in our FAERS-based analysis (ROR: 1506.82), which aligns with published clinical data showing up to 61% of users developing visible skin hyperpigmentation.^18,19 Unlike post-inflammatory pigmentation, this effect appears to be a direct consequence of melanocortin receptor activation, especially MC1R (melanocortin 1 receptor),^20,21 offering a unique perspective on drug-induced pigmentation and highlighting a pathway potentially exploitable for therapeutic repurposing in pigmentary disorders.

In our disproportionality analysis, ribociclib emerged as one of the top-ranking agents associated with hypopigmentation (ROR = 60.20), despite the absence of any pigmentation-related AEs in its product labeling. Ribociclib is a selective cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitor approved for the treatment of hormone receptor-positive, human epidermal growth factor receptor 2-negative metastatic breast cancer.²² The pathogenesis of ribociclib-induced hypopigmentation remains uncertain. Although it does not directly target immune pathways, inhibition of the p16 INK4a/Cyclin D1–CDK4/6–retinoblastoma protein axis may disrupt melanocyte homeostasis. The absence of autoimmune history in reported cases suggests a mechanism distinct from classical vitiligo, though a noncanonical immune contribution cannot be excluded.²³

Our findings are consistent with prior clinical observations.^22,24,25 For example, Alexander et al reported vitiligo-like lesions in patients receiving CDK4/6 inhibitors, particularly ribociclib.²² Vitiligo-like lesions associated with ribociclib are typically resistant to conventional skin-directed therapies, with only mild repigmentation observed in isolated cases treated with ruxolitinib cream.²⁶ The poor treatment response of ribociclib-induced vitiligo-like lesions warrants deeper exploration of their mechanisms and clinical relevance.²⁷ This association raises important questions regarding the underlying mechanisms, including whether it involves immune modulation or melanocyte-targeted cytotoxicity. Further research is needed to elucidate these pathways and potentially inform therapeutic strategies for pigmentary disorders.

In contrast, triamcinolone, a mid-potency corticosteroid, also demonstrated a strong hypopigmentation signal (ROR = 37.20), which is consistent with known effects documented in the prescribing information. Corticosteroids are well-established inhibitors of melanogenesis, acting through suppression of tyrosinase activity, downregulation of microphthalmia-associated transcription factor, and reduced dendritic transport of melanosomes. Chronic use may also induce local skin atrophy and structural disruption, further exacerbating pigmentary loss. In a comparative study, intralesional triamcinolone was associated with hypopigmentation in 80% of patients and skin atrophy in 73.3%, supporting these mechanistic insights.²⁸ The concordance between FAERS data and clinical labeling in triamcinolone underscores the reliability of pharmacovigilance signal detection in capturing expected AEs.

Certain agents, including dupilumab, tretinoin, and clobetasol, exhibit bidirectional pigmentary modulation, manifesting as both hyperpigmentation and hypopigmentation. This observation highlights the complexity of drug-induced pigmentary changes, which are not always unidirectional and may involve intricate immunological, inflammatory, and melanocytic mechanisms—especially in the case of biologics that act through immune pathways. Moreover, factors such as patient ethnicity, underlying dermatologic conditions, and individual immune history may further influence the pigmentary response. Given this complexity, clinicians should be aware of the potential for both pigmentary outcomes when prescribing these agents, and patients should be appropriately counseled regarding these possible effects.

These findings highlight the need to strengthen post-marketing surveillance for agents with high disproportionality signals in the FAERS database, particularly those not currently reflecting pigmentary AEs in their prescribing information. Notable examples include ribociclib and amlodipine, which exhibited significant associations with hyper- or hypopigmentation but lack corresponding label warnings. Furthermore, for drugs frequently prescribed in chronic disease contexts, it is advisable to establish standardized protocols for routine dermatologic monitoring of long-term users, with a focus on early detection and documentation of pigmentary changes. This study also provides a foundation for future large-scale cohort studies or clinical trials to validate the causal relationship between drugs and pigmentary changes. Moreover, the identified agents may offer novel opportunities for drug repurposing, such as utilizing hyperpigmentation-inducing drugs in treating hypopigmentary disorders (eg, vitiligo).

Several limitations of this study should be acknowledged. First, FAERS data are subject to underreporting, reporting bias, and incomplete information, likely leading to an underestimation of actual AE incidence. Second, disproportionality analysis can reveal associations but not establish causality. Third, the absence of detailed clinical information—such as lesion morphology, outcome, or histopathology—limits mechanistic interpretation. Fourth, confounding factors such as polypharmacy and comorbidities may influence the observed signals. For instance, concurrent use of minocycline with photosensitizing agents may intensify hyperpigmentation, while hepatic, endocrine, or chronic inflammatory disorders can independently cause pigmentary changes. Finally, the exclusion of non-cutaneous pigmentary events (eg, ocular or nail involvement) may underestimate the overall pigmentary burden.

Conclusion

This pharmacovigilance analysis of the FAERS database from 2010 to 2024 identified a spectrum of drugs associated with hyperpigmentation, hypopigmentation, or both, some of which are not currently labeled for such risks. Disproportionality signals were particularly strong for some agents, including minocycline, ribociclib, and setmelanotide, underscoring the need for strengthened dermatologic monitoring, especially in long-term therapy. Drugs with bidirectional pigmentary effects—such as dupilumab, hydroxychloroquine, and clobetasol—were also identified in our analysis. Considering the odds ratios and analytical findings, the drugs implicated in either hyperpigmentation or hypopigmentation warrant further investigation to better elucidate their underlying mechanisms and clinical relevance.

Abbreviations

AE, Adverse Event; ADR, Adverse Drug Reaction; CDK4/6, Cyclin Dependent Kinase 4 and 6; CI, Confidence Interval; FAERS, FDA Adverse Event Reporting System; MC1R, Melanocortin-1 Receptor; MedDRA, Medical Dictionary for Regulatory Activities; PT, Preferred Term; ROR, Reporting Odds Ratio.

Data Sharing Statement

The data used in this study were obtained from the publicly accessible FDA Adverse Event Reporting System (FAERS), available at https://www.fda.gov/drugs/questions-and-answers-fdas-adverse-event-reporting-system-faers. All data were fully de-identified by the original data provider. The datasets analyzed during the current study are available from the corresponding author upon reasonable request.

Ethical Considerations

This study is exempt from ethics approval according to Items 1 and 2 of Article 32 of the Measures for Ethical Review of Life Science and Medical Research Involving Human Subjects (China, February 18, 2023). These provisions stipulate that research may be exempt from ethics review when: (1) it uses legally obtained public data or observational data of public behavior without interfering with such behavior; and (2) it uses anonymized information data that does not contain identifiable personal information. As this study analyzed publicly available, de-identified data from the FAERS database, it meets both conditions above and therefore qualifies for exemption from institutional ethics review.

Acknowledgments

The authors would like to thank the US Food and Drug Administration for providing access to the FAERS database, which served as the primary data source for this study. The interpretations, conclusions, and opinions expressed in this work are solely those of the authors and do not represent the views of the FDA.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

There is no funding to report.

Disclosure

The authors report no conflicts of interest in this work.

References

1. Karimaghaei S, Yee J, Elhusseiny AM. Periorbital skin hypopigmentation associated with the use of topical glaucoma medications. Digit J Ophthalmol DJO. 2025;31(1):29–31. doi:10.5693/djo.02.2024.05.001

2. Laroche M-L, Tarbouriech N, Jai T, et al. Economic burden of hospital admissions for adverse drug reactions in France: the IATROSTAT-ECO study. Br J Clin Pharmacol. 2025;91:439–450. doi:10.1111/bcp.16266

3. Nahhas AF, Braunberger TL, Hamzavi IH. An update on drug-induced pigmentation. Am J Clin Dermatol. 2019;20:75–96. doi:10.1007/s40257-018-0393-2

4. Sutaria A, Rawlani S, Sutaria AH. Causes and management of cutaneous adverse drug reactions: a comprehensive review. Cureus. 2024;16:e55318. doi:10.7759/cureus.55318

5. Thawabteh AM, Jibreen A, Karaman D, et al. Skin pigmentation types, causes and treatment-a review. Mol Basel Switz. 2023;28:4839. doi:10.3390/molecules28124839

6. Lane BN, Hamzavi IH, Lim HW, et al. Concurrent management of vitiligo and acquired disorders of hyperpigmentation: a comprehensive literature review and current practice gaps. Int J Dermatol. 2025;64:485–489. doi:10.1111/ijd.17631

7. Solano F, Briganti S, Picardo M, et al. Hypopigmenting agents: an updated review on biological, chemical and clinical aspects. Pigment Cell Res. 2006;19:550–571. doi:10.1111/j.1600-0749.2006.00334.x

8. Fang H, Su Z, Wang Y, et al. Exploring the FDA adverse event reporting system to generate hypotheses for monitoring of disease characteristics. Clin Pharmacol Ther. 2014;95:496–498. doi:10.1038/clpt.2014.17

9. U.S. Food and Drug Administration. [dataset] FDA Adverse Event Reporting System (FAERS). U.S. FDA; 2024.

10. Natale CA, Duperret EK, Zhang J, et al. Sex steroids regulate skin pigmentation through nonclassical membrane-bound receptors. Elife. 2016:5:e15104. doi:10.7554/eLife.15104

11. Giménez García RM, Carrasco Molina S. Drug-induced hyperpigmentation: review and case series. J Am Board Fam Med JABFM. 2019;32:628–638. doi:10.3122/jabfm.2019.04.180212

12. Eisen D, Hakim MDM-IP. Incidence, prevention and management. Drug Saf. 1998;18:431–440. doi:10.2165/00002018-199818060-00004

13. Mouton RW, Jordaan HF, Schneider JW. A new type of minocycline-induced cutaneous hyperpigmentation. Clin Exp Dermatol. 2004;29:8–14. doi:10.1111/j.1365-2230.2004.01421.x

14. Medeiros FA, Walters TR, Kolko M, et al. Phase 3, randomized, 20-month study of bimatoprost implant in open-angle glaucoma and ocular hypertension (ARTEMIS 1). Ophthalmology. 2020;127:1627–1641. doi:10.1016/j.ophtha.2020.06.018

15. Sharpe ED, Reynolds AC, Skuta GL, et al. The clinical impact and incidence of periocular pigmentation associated with either latanoprost or bimatoprost therapy. Curr Eye Res. 2007;32:1037–1043. doi:10.1080/02713680701750625

16. Karslioğlu MZ, Hoşal MB, Tekeli O. Periocular changes in topical bimatoprost and latanoprost use. Turk J Med Sci. 2015;45:925–930.

17. Jha AK, Sinha R, Prasad S, et al. Bimatoprost in periorbital vitiligo: a ray of hope or dilemma. J Eur Acad Dermatol Venereol JEADV. 2016;30:1247–1248. doi:10.1111/jdv.13176

18. Sridharan K, Sivaramakrishnan G. Adverse event profile of setmelanotide in obesity: an integrated assessment and systematic review using disproportionality analysis, case reports and meta-analysis. Expert Opin Drug Saf. 2025;2025:1–10. doi:10.1080/14740338.2025.2465880

19. Haqq AM, Chung WK, Dollfus H, et al. Efficacy and safety of setmelanotide, a melanocortin-4 receptor agonist, in patients with bardet-biedl syndrome and alström syndrome: a multicentre, randomised, double-blind, placebo-controlled, phase 3 trial with an open-label period. Lancet Diabetes Endocrinol. 2022;10:859–868. doi:10.1016/S2213-8587(22)00277-7

20. Wolf Horrell EM, Boulanger MC, D’Orazio JA. Melanocortin 1 receptor: structure, function, and regulation. Front Genet. 2016;7:95. doi:10.3389/fgene.2016.00095

21. Haluzík M, Müllerová D, Sucharda P, et al. Obesity pharmacotherapy – update 2023. Cas Lek Cesk. 2023;162:19–31.

22. Spring LM, Wander SA, Andre F, et al. Cyclin-dependent kinase 4 and 6 inhibitors for hormone receptor-positive breast cancer: past, present, and future. Lancet Lond Engl. 2020;395:817–827. doi:10.1016/S0140-6736(20)30165-3

23. Garutti M, Targato G, Buriolla S, et al. CDK4/6 inhibitors in melanoma: a comprehensive review. Cells. 2021;10:1334. doi:10.3390/cells10061334

24. Kothari R, Darling H, Bhatnagar A, et al. Ribociclib induced vitiligo-like lesions. Indian J Dermatol Venereol Leprol. 2024;91:396–397. doi:10.25259/IJDVL_239_2024

25. Anjaneyan G, Keechilat P, Duraisamy P, et al. Ribociclib-induced extensive vitiligo-like lesions: possible pathomechanisms with clinical, dermoscopic and histological correlation. BMJ Case Rep. 2022:15:e248782. doi:10.1136/bcr-2022-248782

26. Bang AS, Fay CJ, LeBoeuf NR, et al. Multi-center retrospective review of vitiligo-like lesions in breast cancer patients treated with cyclin-dependent kinase 4 and 6 inhibitors. Breast Cancer Res Treat. 2024;204:643–647. doi:10.1007/s10549-023-07217-2

27. Pasqualoni M, Orlandi A, Palazzo A, et al. Case report: vitiligo-like toxicity due to ribociclib during first-line treatment of metastatic breast cancer: two cases of premature interruption of therapy and exceptional response. Front Oncol. 2023:13. doi:10.3389/fonc.2023.1067264

28. Goyal A, Mehta H, Narang T, et al. A Double-Blinded Randomised Control Study to Compare the Effectiveness and Safety of Intralesional Vitamin D3 with Intralesional Triamcinolone and Its Correlation with Tissue Expression of Vitamin D Receptors in Patients with Keloid. Vol. 32. Wound Repair Regen Off Publ Wound Heal Soc Eur Tissue Repair Soc; 2024:696–703. doi:10.1111/wrr.13209