A patient sits in my chair, holds up her phone, and points at the skin beneath her eyes. "Can you remove these?" she asks. The photo shows a dark band that makes her look permanently tired.

I ask her to tilt her chin toward the ceiling light. The band softens. I ask her to tilt it back. It returns. That small experiment tells me more than any product catalogue could, and it is the reason for this post: many under-eye "dark circles" are not a stain but a shadow, and tear trough filler treats only one of them.

Rembrandt understood the difference. In his late self-portraits, the deep hollows around the eyes are built from light and shadow, not from dark paint. A shadow is not a substance. It is light failing to arrive. You cannot scrub it away, and you should not try to bleach it. You change the surface that casts it.

First, decide what you are looking at

Under-eye darkness has several causes. Some people have pigment in the skin itself. Some have thin skin through which blood vessels show. Some have a structural hollow where the lower eyelid meets the cheek, and that groove catches shadow the way a crease in a sheet does.

Filler belongs to the third group. It cannot lighten pigment, and it cannot thicken skin. What it can do is soften a groove so that light falls more evenly across the area.

In my clinic I check this in a minute: light from above, chin up and chin down, and a gentle stretch of the skin. If the darkness moves or fades with the light, the shadow is the problem. If it stays put, no amount of filler will fix it, and I say so.

What the strongest trial shows

The best controlled evidence comes from a multicenter, randomized, evaluator-blinded study of 333 patients with moderate or severe infraorbital hollows (Biesman et al., 2024). Participants received a hyaluronic acid filler designed for the under-eye area, or no treatment.

At three months, 87.4% of treated patients were responders (at least a one-point improvement on a validated hollow scale), compared with 17.7% of the untreated group (p < .001; Biesman et al., 2024). Needle and cannula delivery performed comparably (p = .967).

Patients also reported recovering within 3 to 5 hours of treatment (Biesman et al., 2024). That is a self-reported figure, and it says nothing about how long any bruising lasts in an individual.

How long does it last, and what does it cost in side effects?

In the same trial, 63.5% of patients were still responders at 12 months after the first treatment. After retreatment at month 12, 80.3% were responders at month 18 (Biesman et al., 2024). Filler here is a maintenance decision, not a one-time event.

A single-practice retrospective review of 155 patients treated between 2007 and 2023 found significant improvement persisting at 18 months, with no significant difference in change scores between the 6-, 12-, and 18-month checks (Puyana & Montes, 2025). Those patients received an average of 0.45 mL per hollow through a cannula. It is a useful signal, but it is not a controlled comparison.

On safety, a meta-analysis of 31 reports and 2,556 participants estimated swelling in 19.2% and bruising in 18.4% of patients, with redness at 7.1%, contour irregularity or lumps at 5.3%, and blue discoloration (the Tyndall effect) at 0.9% (Liu et al., 2024). In the randomized trial, 12.7% of patients reported adverse events, mostly mild, with 4.9% in the needle group and 20.9% in the cannula group (Biesman et al., 2024).

The strongest objection

Some will say that a product trial is a marketing exercise, and that a 91% satisfaction rate proves little. That is partly true. The randomized trial evaluated a single manufacturer's product, so its numbers may not transfer to every filler on the market. The meta-analysis pooled reports that were mostly uncontrolled, and its search ended in September 2022 (Liu et al., 2024).

But the direction is consistent across a randomized trial, a pooled analysis, and a long-run clinical series. And the pooled analysis reports satisfaction of 91.0% (95% CI 84.9 to 95.7) alongside a low rate of complications (Liu et al., 2024). What these papers cannot tell us is how often rare, serious events occur. Reviews of this size are poorly equipped to capture them, which is why who injects, and how, matters as much as what is injected.

The Clinical Bottom Line

Before you book, ask three things. First, is my darkness a hollow, pigment, or thin skin? Second, if it is a hollow, which product and which technique (needle or cannula) will you use, and why? Third, what is the plan for maintenance, since the trial data suggest retreatment is part of the picture?

If the answer to the first question is pigment or visible vessels, filler is the wrong tool, and a good clinician will tell you so before taking your money. I’ll be sharing my case reports in the near future to share how I’ve evaluated different patients and what the results look like and how sometimes consultants in Korea can be wrong about the product choice.

A shadow is not a stain. Treat the one you actually have.

For clinical assessments and consultation in Seoul: itsdrbock.com

References

Biesman, B. S., Green, J. B., George, R., Jacob, C., Palm, M., Jones, D. H., Grunebaum, L., Beer, K., Cho, Y., Joseph, J. H., Almegård, B., Weinberg, F., & Bromée, T. (2024). A multicenter, randomized, evaluator-blinded study to examine the safety and effectiveness of hyaluronic acid filler in the correction of infraorbital hollows. Aesthetic Surgery Journal, 44(9), 1001–1013. https://doi.org/10.1093/asj/sjae073

Liu, X., Gao, Y., Ma, J., & Li, J. (2024). The efficacy and safety of hyaluronic acid injection in tear trough deformity: A systematic review and meta-analysis. Aesthetic Plastic Surgery, 48(3), 478–490. https://doi.org/10.1007/s00266-023-03613-7

Puyana, C., & Montes, J. R. (2025). Long-term effects of tear trough hyaluronic acid filler: A retrospective study. Journal of Clinical and Aesthetic Dermatology, 18(11), 44–47. https://jcadonline.com/long-term-effects-of-tear-trough-hyaluronic-acid-filler-a-retrospective-study/