Biofouling Mitigation in Dermal Sensors

Biofouling Mitigation in Dermal Sensors

When a glucose sensor sits in the skin itself instead of beneath it, the body's reaction to it is smaller, and the problems that remain look different. The Biolinq Shine, which the FDA authorized in September 2025, places seven microelectrodes in the papillary dermis, the uppermost layer of the dermis, and is worn for up to 5 days [1]. Wearables that read sweat never make a wound at all. Their trouble is a sample contaminated by whatever is already on the skin [2], and an older device that pulled glucose out through the skin ran into irritation bad enough that users wore it less [3].

Our article on biofouling mitigation strategies for dermal sensors explains the coatings and drug-release designs built for sensors that stay in longer. This one stays with sensors that live in or on the skin, where the main defense we found is a short wear period.

Three places a sensor can sit

Conventional CGMs use an indwelling sensor needle to measure glucose in interstitial fluid, the fluid that bathes the cells [4]. Microneedle sensors don't go as deep. A prototype from Sweden's KTH Royal Institute of Technology put its electrodes inside a single hollow microneedle that reached the dermis, and the part in the skin was 50 times smaller than that of commercial products [5]. An ArKal Medical prototype took another route: hollow microneedles in the epidermis, the outer layer of skin, let glucose diffuse to a sensor that stayed outside the body [6].

Surface devices don't enter tissue. Sweat sensors read fluid the body has already released. The GlucoWatch Biographer, described by its maker Cygnus in 2000, extracted glucose through intact skin and gave readings for up to 12 hours at a time [7].

So "fouling" changes with depth. In the dermis it is the familiar buildup of proteins and immune cells, around a much smaller wound. On the surface it becomes a contaminated sample, plus the skin's reaction to the device itself.

A smaller wound that closes fast

The KTH team reported that the smaller footprint reduces tissue trauma and was virtually painless to insert, and in human tests their device tracked blood glucose with about a 10-minute delay [5]. Skin also repairs itself quickly. In a Georgia Tech study in people, skin pierced by microneedles regained its barrier within 2 hours when left uncovered, while covered (occluded) sites took 3 to 40 hours, longer for bigger and more numerous needles [8]. That second figure is closer to a wearable's situation, since a patch sensor sits under an adhesive cover for as long as it is worn [1].

What we didn't find was long human data. One microneedle array from Imperial College London was tested in a single healthy volunteer for 3 to 6 hours [9]. The ArKal device ran for up to 72 hours in 10 people with diabetes. Its mean absolute relative difference (MARD, the average gap between sensor and reference readings) was 15%, and the authors described it as well tolerated [6]. A University of California San Diego team paired a disposable microneedle array with reusable electronics and tested it in volunteers going about daily activities [10]. Our page on microneedle array sensor innovation covers how these arrays are built.

Disposability is the mitigation. Surface properties shape the foreign body reaction most in the first two to four weeks after implantation, according to a 2008 review by Anderson and colleagues [11]. A patch replaced within days never gets far into that window, which sidesteps the long-term problem without solving it.

What the FDA review of the Biolinq Shine shows

Of the intradermal sensors we read about, the Biolinq Shine has the most human data, because it went through FDA review. The De Novo authorization covers prescription use in adults 22 and older who don't use insulin [1], and in September 2026 the FDA cleared over-the-counter labeling for the same device [12]. Its electrodes use glucose oxidase, an enzyme whose reaction produces a current proportional to the glucose in interstitial fluid, and the patch shows colored lights instead of numbers [1].

In the pivotal study, 96.3% of the primary wearables were still working on day 1, and 84.7% lasted the full 5 days [1]. Here's what surprised us: agreement with blood tests held steady across those days. When the patch showed its blue in-range light, a venous reference sample fell in the same range 80.3% of the time on day 1, 79.0% on day 3 and 80.0% on day 5 [1]. Color-band agreement is a coarser measure than the numeric accuracy reported for CGMs. The review also doesn't say why some sensors stopped early, so we don't read that drop-off as a biofouling rate.

Biocompatibility testing on the parts that touch the body included sensitization, irritation and implantation tests, and the review concludes that the device meets ISO 10993-1, the international standard for biological evaluation of medical devices [1]. The labeling limits wear to the inner forearm and says not to place the sensor on irritated or damaged skin [1].

On the skin, fouling means a contaminated sample

A sensor that never enters the body escapes the foreign body reaction, but the skin surface brings its own contamination. In 23 studies of seven people with diabetes, forearm sweat tracked blood glucose closely, but only when it was collected in a way that kept out glucose already sitting on the skin [2]. A 2019 review, whose co-authors included a researcher from Abbott Diabetes Care, argued that the fluid itself is often the biggest obstacle: tiny sample volumes and changing secretion rates, shifting pH and salinity, and breakdown of the molecule being measured [4].

Sweat also dilutes what it carries, and enzymatic sweat sensors are sensitive to flow [13]. Some designs stimulate sweat with a drug delivered into the skin, but those stimulants act on more than sweat glands; a University of Cincinnati team warned that placing them directly on skin risks uncontrolled transport into the body and contamination of the sweat sample [14].

Skin reactions: the problem every wearable shares

The GlucoWatch Biographer shows what happens when the skin objects to the measurement itself. In Cygnus's summary of its trials, skin irritation, meaning redness and swelling, was absent or mild in more than 87% of users and cleared within several days without treatment [7]. Longer use told a harsher story. In a six-month randomized trial of 200 children and teens published in 2005, use of the GlucoWatch G2 Biographer fell from 2.1 to 1.5 times a week, and skin irritation was the most common reason given for wearing it less (76%) [3].

Adhesives cause a separate set of reactions that any taped-on device can trigger, microneedle patches included. European dermatologists identified isobornyl acrylate (IBOA) as a major contact allergen in glucose sensors and insulin pumps [15]. In one Belgian dermatology department, 12 children were patch-tested after reactions to glucose sensors or infusion sets; 10 reacted to the adhesive part of an earlier FreeStyle Libre sensor, and 10 were sensitized to IBOA [16]. Not every reaction is an allergy. In a series of 52 patients from the same department, 17 tested negative to both IBOA and the device adhesive, which the authors attributed to irritant contact dermatitis or an allergen nobody had tested for yet [17].

Those Belgian studies predate Abbott's current US sensors; as of October 2026, those are the FreeStyle Libre 2 Plus and 3 Plus [18]. Our article on adhesive chemistry and dermatological issues in CGMs goes further into the chemistry. Whether to keep wearing a device that irritates your skin is a decision to make with your care team.

What we still don't know

In the sources we read, human wear of intradermal glucose sensors stops at 5 days [1]. That is long enough to show a patch can hold steady agreement through its wear, and too short to say how the dermis would respond to a sensor left in for weeks. The sweat-to-blood link rests on small studies, such as the seven-person forearm series [2]. The studies we read don't settle whether sitting shallower buys a gentler reaction or only a different one.