Refrigeration and Skincare: What the Research Actually Shows

Does refrigerating skincare extend its shelf life?

For the chemistry inside the bottle, yes, and the measurements are not subtle. The clearest cosmetic example is a 2022 study in which cosmetic creams containing a grape-seed polyphenol extract were split across three storage temperatures, 4°C, 25°C and 50°C, then tracked for four months and followed past eight. Total phenolic content held up best in the refrigerated jars and fell hardest in the hot ones, and the authors state it in one line: the highest stability was achieved at 4°C and the least at 50°C ^[5]. Two details keep that honest: the hot arm at 50°C works as a stress condition, not a stand-in for a warm shelf, and the measurement, phenolic content in milligrams of gallic acid equivalent per milliliter, is a chemical assay of the extract instead of a percentage of a product's advertised strength ^[5].

A polyunsaturated oil gives the longer timeline. Hemp seed oil was stored for 270 days in four packaging materials, one set at 10°C in the dark and the other at 25°C under diffused light. The refrigerated samples held peroxide value below 10 milliequivalents of oxygen per kilogram in every glass packaging for the full 270 days, with only the oxygen-permeable polypropylene creeping past that limit at the very end, while room-temperature polypropylene vials reached 40 by day 150 ^[11]. Higher temperature accelerated the free-radical reaction that oxygen starts, and light increased oxidation as well ^[11]. That second clause is the catch: the arms differed in light as well as temperature, so this is a cool-and-dark result, not a clean temperature-only one, and peroxide value is an oxidation marker for edible oils, never an assay of cosmetic potency.

The study that measured real products on real shelves

The most useful evidence in this literature does not come from a model solution. In 2021 a research group bought 12 commercial cosmetics, tracked 16 retinoid derivatives in them across six months of long-term and accelerated stability testing, and added a one-week photostability study. Almost every product showed some decline: after six months the loss ran from 0% to 80% at 25°C and from 40% to 100% at 40°C, following first-order kinetics in each case ^[3].

Three details belong next to that number. The spread is enormous, and the authors attribute it to formulation. Light mattered more than heat in their hands, which cuts against a temperature-only story and is worth printing anyway. And their own conclusion is methodological: topical formulations need real-time stability studies instead of the accelerated tests the industry more often runs ^[3]. Wild Ice sells no retinoids, and none of this is a verdict on retinol. What it establishes is narrower and more useful: actives in real, finished, commercially sold cosmetics measurably decline over months of ordinary storage, and they decline faster warm.

"Shelf life" is doing two jobs at once

Separating those two jobs prevents most of the confusion here. The FDA uses shelf life to mean the length of time a product can be expected to look and act as expected and to stay safe for use, and no US law requires cosmetics to carry an expiration date at all, so the window is the manufacturer's own judgment ^[4]. The studies above measure something narrower: how much of one specific fragile molecule is still present. A product can sit comfortably inside its usability window while the active you paid for has quietly thinned out.

That distinction is why every number in this article is a question of strength and value, and none of them a hazard warning. The independent US cosmetic-safety panel concluded that ascorbic acid and its ascorbate salts are safe as used in cosmetics ^[23], and the FDA's cautions about older cosmetics center on quality and hygiene rather than on actives turning toxic ^[4]. Degraded is not the same as dangerous, and that holds through every section below, including the microbiology. Reading the visible cues is a separate skill, with its own checklist in seven signs your skincare is no longer fresh.

What did the stability studies actually measure?

Almost nobody asks this, and it is where the interesting part of the answer lives. Stability testing is not a single test but a matrix of defined temperatures, humidities and durations, and the pharmaceutical industry's version of that matrix is written into a guideline called ICH Q1A(R2) ^[1].

Note what happens in the second row. For a product intended to be stored in a refrigerator, the guideline's accelerated condition, the stress test, is ordinary room temperature ^[1]. Room temperature is not a neutral baseline in this framework. For anything meant to live cold, room temperature is the accelerating condition, and that statement comes from a regulator, not a marketer.

Two boundaries come with that table. ICH is a pharmaceutical guideline and cosmetics are not required to follow it; the FDA states that shelf life is the manufacturer's responsibility, and that it is aware some companies test products by exposing samples to different temperatures and humidity levels ^[4]. And Wild Ice has not run ICH stability testing on anything. Those conditions are the discipline the cosmetics industry borrows from, and the reason 25°C and 40°C recur in every study quoted here.

Tap to open full size

Cosmetic preservation is tested at room temperature, on purpose

Preservative systems get a separate test, and its conditions are quietly relevant. A challenge test determines the efficacy and stability of the preservative system over time, and in the standard protocol the containers are protected from light and incubated at room temperature for 28 days ^[19]. A preservative system is therefore validated on the assumption that the product will live warm. Whatever cool storage adds, it is added on top of a system already asked to work without it.

Why anyone trusts a hot test to predict a cool shelf

The bridge between a three-week oven and a two-year shelf is the Arrhenius relationship, which ties degradation rate to temperature tightly enough that its logarithmic form is a straight line, so a rate measured hot can be extrapolated down to mild conditions ^[2]. The same work adds the corollary that matters for fragile ingredients: the lower a molecule's activation energy, the less stable it is ^[2]. A printed shelf life is usually a prediction made with that equation: the product is aged on purpose in an oven, so that a year of shelf aging is over in a matter of weeks ^[24]. We run the same dial the other way. Turned down to about 4°C, it makes the reactions the oven hurries crawl instead, and nothing in the equation exempts a serum from either direction. The principle itself, how heat and time drain what a serum can do, has its own explainer.

The catch is in the working conditions. Accelerated predictive stability studies run over three to four weeks at 50°C to 80°C ^[2], and that extrapolation is only as good as its central assumption, which is that the same degradation pathway operates at 70°C as at 20°C. Sometimes it does not, and the limits section below collects what follows.

The studies, one by one, and what each one does not show

Below is the evidence this article rests on, at study level. The load-bearing column is the last one, because a study quoted without its limits is decoration rather than evidence. Two rules govern every number in the table: each figure travels with the material and the conditions it was measured under, in the same row, and no percentage anywhere in this article is attached to a Wild Ice product, since we have no published stability assay on any of them.

Read as a set, the pattern is consistent: cool storage slowed degradation in every material anyone measured it in, the loss followed first-order kinetics almost everywhere, and the size of the effect varied with the formula, the marker and the molecule. Read as proof of a shelf life for a particular serum, the set says nothing. That consistency is what our whole approach to preserving skincare with cold is built on. The timeline view of this same evidence, ingredient by ingredient over weeks, is assembled in how fast skincare loses its potency.

Does cold slow bacterial growth?

Yes, substantially, and the plainest statement of it sits in food-safety guidance. The FDA calls chilling stored foods to proper temperatures one of the best ways to slow the growth of dangerous bacteria, notes that at room temperature the numbers of bacteria that cause foodborne sickness can double every 20 minutes, and gives the recommended refrigerator temperature as 40°F or below, about 4°C ^[17].

The microbiology behind that is well-mapped. Mesophiles, the group with optimal growth temperatures from about room temperature to about 45°C, include normal human microbiota and organisms such as E. coli and Salmonella, so refrigeration puts them well below their preferred range ^[16]. What most brand copy leaves out is the other half. Psychrotrophs prefer cooler environments, from a high of 25°C down to refrigeration temperature around 4°C, and they are the organisms responsible for spoiling refrigerated food ^[16]. Listeria monocytogenes is the textbook example of one that multiplies at 4 to 10°C ^[16], and the FDA notes the same in its fridge guidance ^[17].

So the accurate sentence is that refrigeration is a powerful hurdle rather than a sterilizing step. Cold slows growth sharply and stops it for many organisms, while some keep going slowly. That is a statement about growth rates in food and clinical microbiology, and it carries no suggestion that anyone's cosmetic product is contaminated or that a warm serum is a hazard. Those are different claims, and only the first has evidence behind it.

Does refrigeration replace preservatives?

No, and the literature is unusually direct about why. A review of cosmetic contamination and preservation states that cosmetics with high water content are at risk of being contaminated by microorganisms that can alter the product or pose a health risk, that manufacturers add preservatives to avoid this, and then draws the line that governs this entire article: products with low water activity are self-preserved, but many cosmetic products such as creams, lotions, shampoos, conditioners and liquid soaps all have high water activity, and consequently chemical preservation is necessary ^[20].

Water activity is the measure underneath that, defined as how much of a product's water content is available for microbial use ^[20]. The preservation literature supplies the thresholds: a water activity value of 0.8 can guarantee microbiological stability without preservatives in the formulas tested, and microorganisms cannot proliferate or survive in a cosmetic formulation at a pH below 4 or above 10 ^[19]. Its name for stacking several of these protections is hurdle technology, the intelligent combination of the several factors that prevent the development of microorganisms ^[19]. Notice which factors that review enumerates for cosmetics: water activity, pH, and the preservative system. Refrigeration is not on its list. Cold is a recognized way to slow both the chemistry and the biology, and it sits alongside those formulation hurdles rather than standing in for any of them.

Tap to open full size

Heat works against the preservative system too, which is the least intuitive part of the picture and the most useful. The FDA's storage advice says heat can make preservatives break down and cause bacteria and fungi to grow faster, and elsewhere it notes that preservatives can break down over time, allowing bacteria and fungi to grow ^[4]. Its guidance on microbiological safety repeats the point from the other side: store products as instructed, because some microorganisms grow faster in warm or damp conditions and the preservatives in the product may break down ^[21]. Cool storage and a preservative system are therefore allies. One protects the other.

It also helps to see how narrow refrigeration's jurisdiction is. The FDA lists the ways cosmetics become contaminated: contaminated raw materials, water or other ingredients; poor manufacturing conditions; an ineffective preservation system; packaging that does not adequately protect the product; poor shipping or storage conditions; and consumer use, such as dipping fingers into the product ^[21]. Storage temperature is one item on a six-item list, and consumer use sits on the same list, which is why a product open on a bathroom shelf needs a preservative system regardless of how cold it was kept beforehand.

Is fresh production automatically clean production?

One study is worth reading precisely because it complicates the small-batch story instead of flattering it. Researchers in a national public-health pilot collected 120 customized cosmetic samples directly from shops and found aerobic microbes in 49 of them (40.8%), with counts from 10 to 2,830 colony-forming units per gram or milliliter; products transferred to new containers during production showed the higher counts, heat treatment reduced them, and the authors concluded that the preservation system of on-the-spot customized cosmetics cannot be assured, because they are made on site and sent directly to consumers ^[22].

That study surveyed mix-to-order retail cosmetics in one national pilot, and it is not a finding about small-batch manufacturers as a class or about anyone's brand. Its value here is the discipline it enforces on our own argument: freshness is a claim about actives, microbiological safety is a separate discipline handled by formulation and process control, and cold contributes to both without being sufficient for either. The wider point is best stated in the negative. Preservatives address one of the pathways by which a product deteriorates, not all of them, and the chemistry that drains an active was never their job. Which formulas can safely go without a preservative system, and which have to carry one, is a different argument, and it gets made in whether natural skincare needs preservatives, and which of ours name them.

What the evidence says, active by active

Four families of fragile active carry most of the weight in the formulas we make, and each has its own literature with its own gaps.

Polyunsaturated botanical oils

This is the best-quantified family, because food science has been measuring oil oxidation for a century. The review literature states the Q10 rule of thumb cleanly: the oxidation rate of oils is generally doubled for each 10°C increase in temperature ^[10]. It also explains why polyunsaturated oils sit at the fragile end, since in a relative-rate comparison at 100°C the C18 series runs stearic acid at 1, oleic at 10, linoleic at 100 and linolenic at 150 ^[10]. Those are relative reactivities at a high temperature, so the ranking is real while the multiplier forecasts nothing about a bottle on a shelf.

The levers come from the same review: lowering storage temperature is described as an excellent way to control lipid oxidation, oxidation cannot happen without oxygen at all, and losing an oil's native antioxidants such as tocopherols compromises its oxidative stability ^[10]. Rosehip, sea buckthorn, cranberry seed and rice bran oils, which we use across Cryo-C and Berry Boreal Bod, belong to this class, and vitamin E is the in-formula antioxidant that literature describes as protective from the inside.

Resveratrol

The resveratrol evidence is real and thinner than the oils'. trans-Resveratrol is stable in acidic conditions, its degradation begins to rise exponentially above pH 6.8, and increased temperature likewise compromises its stability ^[13]. No fetched study gives it a half-life at storage temperatures, so any ranking that places resveratrol among the fastest to fade is qualitative by necessity.

The same paper carries the best methodological warning in this literature: a common assay, UV/VIS spectroscopy, returned falsely higher trans-resveratrol concentrations under exactly the conditions in which the molecule was unstable, meaning alkaline pH, light and increased temperature ^[13]. Choose the wrong instrument and a degrading sample can read as a healthy one.

The vitamin C ester

Ascorbyl tetraisopalmitate, the oil-soluble vitamin C in Cryo-C and Berry Boreal Bod, appears in the literature under its synonym tetrahexyldecyl ascorbate: a lipid-soluble ascorbic-acid precursor esterified with a branched-chain fatty acid, reported to have improved stability and an improved ability to penetrate the lipophilic stratum corneum ^[9]. The same paper carries the other half. Under a singlet-oxygen oxidative-stress assay the ester degraded completely by six minutes without a co-antioxidant present, and only 25% in ten minutes with one ^[9]. A minutes-long stress assay is nothing like shelf storage, so nobody should read it as a bottle's lifespan. What it supports is the design choice: pair the ester with another antioxidant, and keep it cool and dark.

For the vitamin C class more broadly, free ascorbic acid is the least stable form and the best documented. It is reversibly oxidised to dehydroascorbic acid on exposure to light, heat, transition metal ions and alkaline pH, then irreversibly hydrolyses to 2,3-diketogulonic acid, and its degradation is significantly reduced at 4 to 10°C ^[6]. The most quotable number in this literature is also the easiest to abuse, so it belongs with its matrix attached: a fruit juice, and a 10°C step that roughly doubles the loss, echoing the oils' rule of thumb without proving a constant.

Peptides

Peptide-bond hydrolysis follows a first-order rate law, and its rate increases exponentially with rising temperature in aqueous solution ^[14]. That study ran at 100 to 220°C in a geochemistry laboratory, so it establishes the mechanism and the shape and nothing about a shelf. It does answer the question people ask most: no temperature exists at which peptides suddenly fail, and the rate climbs with heat.

The pharmaceutical literature adds the practice. Hydrolysis is one of the main degradation pathways of peptides, it is acid and base catalysed and strongly pH-dependent, peptides are often unstable in aqueous solutions, and aqueous injection peptides are often stored at low temperatures and protected from light to minimize degradation ^[15]. Cold storage of water-based peptide products is established pharmaceutical practice, and none of it means a water-based peptide serum is a mistake. Where an active needs water to work, formulators answer with a gentle pH, stabilized peptide forms, and cool, fresh storage, which is the position 7 Winters occupies with its Tripeptide-29.

The ranking, and the study that does not exist

Put those four families in order of how fast they fade under heat and air and you get the order the pillar's card lists: the oils, then resveratrol, then the ester, then a peptide such as Tripeptide-29. The evidence above supports it: the oils lead on relative reactivity and measured peroxide climb ^[10][11], resveratrol is genuinely fragile with no measured storage half-life ^[13], the ester is more stable than free ascorbic acid while still degrading under oxidative stress ^[6][9], and peptide hydrolysis proceeds by a slower route at storage temperatures ^[14][15].

Here is what does not exist anywhere in those sources. No study puts a polyunsaturated oil, resveratrol, an ascorbyl ester and a peptide into one formula and measures all four under one protocol. The ordering is a reasoned assembly across separate literatures with different materials, markers and temperature ranges, and nothing firmer. That is also why the modelled card on our freshness pillar shows relative shapes and prints no numbers on its axes.

Tap to open full size

Built from the twenty-three sources listed below, among them ICH Q1A (2003), González-González 2023, Temova Rakuša 2021, Yin 2022 and Swindell 2021; last checked 2026-09.

Where this evidence stops

Five limits sit on the argument above, and each one changes how the evidence should be used.

1. The material is usually not a serum. Guava juice, rosehip nectar, hot-compressed water, olive oil, hemp seed oil and a compounded chlorhexidine formulation are the matrices behind most of the clean numbers here ^{[6][7][8][11][12][2]}. Only two studies measured a cosmetic, finished commercial products in one case and cosmetic creams in the other ^[3][5], which is why those two carry most of the weight while the rest support the direction rather than the magnitude.
2. Much of it ran far hotter than any storage, and accelerated testing has a ceiling. Half-lives measured at 70 to 95°C in minutes ^[7], peptide kinetics at 100 to 220°C ^[14] and a minutes-long oxidative assay ^[9] describe the shape of degradation and say nothing about a product's life. Accelerated predictive stability itself runs at 50 to 80°C and assumes one pathway operates hot and cool ^[2], and the olive-oil team warned in print that above 60°C vegetable oils develop rancidity levels not relevant to normal storage ^[12]. Fast answers are cheaper and less trustworthy.
3. The marker you choose changes the number. In the same olive-oil model, estimated shelf life at 60°C came out as 32 days on one quality indicator and 4 days on another ^[12]. Resveratrol read falsely high on a UV/VIS assay under the very conditions that destabilized it ^[13]. A published shelf life is a measurement of one chosen marker, not a property of the product.
4. The surrounding formula moves everything. Activation energy for vitamin C degradation shifted from 15.77 to 47.53 kJ/mol across a pH range ^[8], and activation energies for lipid oxidation span roughly 20 to 150 kJ/mol ^[12]. That range is the sourced reason no single Q10 multiplier applies to every ingredient, and why the doubling-per-10°C figure belongs to oils as a rule of thumb, never as a universal constant.
5. Temperature is not the only lever, and sometimes not the largest. In the finished-cosmetics study, light degradation was more pronounced than temperature-induced degradation ^[3], and the nine-month oil study varied temperature and light together ^[11]. Cool matters. Dark matters too, and the evidence says so.

One more limit is ours. Wild Ice has no published stability assay on Cryo-C, 7 Winters, Berry Boreal Bod or anything else, so no percentage in this article is attached to a Wild Ice product, and none should be without testing behind it. These studies model storage conditions in general, which is why a product's history before it reaches you is its own question with its own evidence, covered in whether skincare degrades before you buy it. Nor do they settle whether your own bottle belongs in the fridge, since some facial oils cloud or thicken when chilled and the answer varies by product; that practical decision has its own guide, how to store your skincare.

Medicine already does this, and the analogy has limits

Temperature-controlled handling of fragile molecules is a long-established practice elsewhere. Vaccine handling runs on what the CDC defines as a temperature-controlled supply chain, covering all vaccine-related equipment and procedures from cold storage at the manufacturing plant through transport and provider storage to administration, with refrigerators maintaining 2°C to 8°C ^[18].

The boundary matters more than the parallel. Vaccines are biologics governed by a regulated clinical requirement, cosmetics are not, and nothing in this article makes any claim about temperature after a package leaves us. What the medical precedent legitimately shows is that cool storage is a recognized stability strategy for fragile molecules, and the peptide-formulation literature records the same habit: aqueous peptide injectables are routinely stored cold and protected from light ^[15]. A serum is not a vaccine and the stakes are nowhere near comparable. The chemistry of keeping a fragile molecule intact is the same either way.

How we read this evidence in the studio

I formulate the products this evidence bears on, so I will be direct about what we take from it and what we do not.

We take the direction and the size of the temperature effect seriously, because it is consistent across every material anyone has measured. Wild Ice formulas are made fresh in small weekly batches and kept refrigerated at about 4°C in the studio from the moment a batch is formulated until the day it ships, which keeps the fragile actives out of the warm conditions these studies keep measuring. We track cumulative thermal exposure with a degree-day measure we call Heat Age, which keeps a running total of the heat a product has absorbed, so two bottles of the same age can be compared by what they have been through.

Reading the stability literature this closely also shows whose questions it answers. Its working definition of stable belongs to the room-temperature shelf: what its protocols ask of a cosmetic is whether it will come through weeks at 40°C and hotter ^[24] and still, years later, look and act as expected ^[4]. A formula built to pass that test is shaped by it. The stability risk assessment flags the oxidizable oils before anything else ^[24], so the polyunsaturated oils, the fastest of the four families above to oxidize ^[10], are the likeliest thing such a formula leaves out or buries in stabilizers, and every other fragile active gets weighed the same way, by what it can sit through. That is a logistics decision, and the customer pays for it twice, once at the register and once in whatever is missing from the bottle by the time it is opened. Keeping a batch cold from the day it is formulated lets us make the other choice: the polyunsaturated oils and the vitamin C ester at full strength in Cryo-C, resveratrol and Tripeptide-29 in the water base of 7 Winters, each there for what it does. Alongside the four-in-one study that does not exist, there is a second I keep looking for and have not found, the one that asks how much better a serum could be if it never sat warm. The nearest thing to it is what we run on ourselves every week.

Our own catalog makes the water-activity distinction concrete. Cryo-C is an anhydrous oil serum, which puts it in the low-water-activity, self-preserving category the preservation literature describes ^[20]; it carries the vitamin C ester and polyunsaturated botanical oils, two of the four fragile families above, plus vitamin E as the in-formula antioxidant. D.B.T. Baba Yaga is a water-containing cream, carrying its palmitoyl dipeptides and ceramides in a rose-distillate base, and its ingredient list names three preservatives, benzyl alcohol, salicylic acid and sorbic acid, because a high-water cosmetic needs that protection built in ^[20]. That is the evidence in this article applied to two different formulas, neither of which is built to wait years on a warm shelf, so neither carries a preservative system engineered for that. Cold protects the actives that preservatives were never designed to protect, and it does not do the preservative's job.

The home habit that follows depends on which formula you are holding. Thin water-based products, toners and lightweight treatments, do best refrigerated after opening for freshness. A waterless oil serum like Cryo-C and a rich cream like D.B.T. are happier in a cool, dark cabinet, where the fridge is optional and stiffens the texture more than it buys you, since oils and heavy emulsions cloud and thicken when chilled. Either way, a fresh bottle used up beats a full one kept.

Reading the evidence this way is what produced our whole approach to preserving skincare with cold, instead of pushing shelf stability with formulation chemistry alone.