"How long does it last" bundles two different questions. One is whether the fat cells the drug destroyed can regrow. The other is whether the visible result under your chin holds. They have different answers and different quality of evidence behind them.
The first is a biology question, and the biology is reasonably clear. The second is a follow-up question, and the follow-up runs to three years, in a selected group, measured by eye.
The trials that supported approval followed people for 24 weeks after their last treatment. A separate study then re-enrolled participants and followed them, without further treatment, for three more years across 13 US and 2 Canadian sites. Of 257 people screened, 224 enrolled and 200 completed.
Its primary measure was whether a clinician-rated one-grade improvement was still there. The results, drug against placebo: year 1, 86.4 against 56.8 per cent. Year 2, 90.6 against 73.8 per cent. Year 3, 82.4 against 65.0 per cent. On the stricter two-grade clinician measure, among people who had achieved it originally: 75.0 against 18.2 per cent at year 1, 72.5 against 33.3 per cent at year 2, and 65.8 against 33.3 per cent at year 3.
Satisfaction held up too. Of treated patients who were satisfied at 12 weeks, 74 per cent were still satisfied at year 3. Skin laxity was unchanged or improved at year 3 in 75 per cent of treated patients and 74 per cent of placebo patients.
No new treatment-related side effects appeared over the three years, and every treatment-area effect reported had first appeared during the original trials.
The study was sponsored by Allergan, and Allergan performed the data analyses: One author was an employee of AbbVie, and the others were company investigators or consultants. That does not make the numbers wrong. It does mean they are not independent, and the paper's own limitations section names selection bias, attrition bias and small numbers in the two-grade subgroups.
Three limits matter more than the headline.
Only responders were let in: You had to have improved by 12 weeks after your last trial treatment to enrol. So the study answers "if it worked for you, does it hold?" — not "will it work and hold for me?". Across the original trials, 66 to 70 per cent achieved a one-grade improvement on the combined clinician and patient scales, and only 13 to 19 per cent achieved a two-grade improvement.
Nothing was measured objectively: The trials did use MRI, but only in a subgroup of 449 people and only at 12 weeks, where 43 per cent of treated patients achieved at least a 10 per cent reduction in fat volume against 5 per cent on placebo. After 12 weeks, no MRI, no ultrasound and no 3D imaging was done anywhere in the programme. The follow-up authors list this as a limitation themselves. The three-year evidence is people looking at necks and assigning grades.
The placebo group held its result too: Between 56.8 and 73.8 per cent of people who had "responded" to placebo injections were still rated as responders years later. Since placebo injections do not destroy fat, that number is measuring something else: raters drifting, people's own memory of their baseline, ordinary variation in how a neck looks. Anywhere between half and two thirds of the maintained rating in the treated group could be the same effect. The gap between the groups is the drug's contribution, and at year 3 that gap is 17 percentage points.
The pattern is sharpest on the patients' own strict ratings. Maintenance of a two-grade patient-rated improvement was 59.0 against 63.6 per cent at year 1, 57.5 against 60.0 at year 2 and 59.5 against 70.0 at year 3. No separation at all, and the placebo group numerically ahead. The authors say plainly that there is a substantial placebo effect in patients' own ratings.
A review article published in 2016, written by authors affiliated with the developer, states that treatment response is maintained "up to 4 years" after the last treatment. That sentence is the source of nearly every four-year claim online.
Behind it is a long-term follow-up of an early-phase group, planned to run five years, first presented at a dermatologic surgery meeting in 2011 with two years of data and announced in a company press release. There is no peer-reviewed publication of four-year follow-up data.
That is not an accusation that the four-year data are wrong. It is a statement about what can be checked. A conference presentation and a press release are not a paper you can read, and a claim you cannot inspect should not carry the same weight as one you can. Three years is the number a page like this can stand behind.
There is also a shorter but useful study: a 12-month open-label study in 165 people, with no control group, in which 90.4 per cent of clinician-rated responders and 80.7 per cent of patient-rated responders still had their improvement at 12 months. Open-label and uncontrolled, so it cannot separate the drug from everything else, but it points the same direction over the first year.
The label's whole mechanism section is one sentence: the drug physically destroys the cell membrane, causing the cell to burst. Tissue examined 28 days after injection shows shrunken fat lobules, thickened fibrous partitions and new small blood vessels, with all changes confined to the fat layer.
A cell that has burst is gone: In adults, fat-cell number is broadly stable, and weight change is mostly cells getting bigger or smaller rather than new cells appearing. That is why "the cells do not come back" is a reasonable statement.
Here is what it does not mean. It does not mean the label claims permanence — the word does not appear in it, and the approved indication is improvement in the appearance of fullness, not fat removal. It does not mean every fat cell in the treated area was destroyed; the average trial patient received 25.3 mL in total out of a permitted maximum of 60 mL, and fewer than half achieved even a 10 per cent measured reduction in fat volume. And it does not mean the area is frozen in place. Cells you still have can grow.
The remaining fat cells in a treated area can enlarge, in the same way fat cells anywhere enlarge with weight gain. This is physiological reasoning, not a documented finding: no study has measured fat-cell size in a Kybella-treated neck after subsequent weight gain.
The nearest hard evidence comes from a different procedure. The only randomised trial of whether fat returns after it is removed studied small-volume liposuction in 32 non-obese women. At one year, overall body fat had been restored, and the pattern had shifted — thigh fat stayed reduced while abdominal fat came back. A much larger single-surgeon photographic series of 301 liposuction procedures concluded the opposite, that there was no evidence of fat regrowth. Neither studied injection lipolysis, and the two disagree with each other, so anyone telling you fat definitely does or definitely does not come back somewhere else is extrapolating.
The practical version: destroying fat cells under the chin does not change what your body does with surplus calories.
Two things in the data are worth knowing before you start.
Dose and sessions: The label permits up to six treatments at least a month apart. In the trials, the median number of sessions to a combined clinician-and-patient response was 3, and 59 per cent of subjects went through all six. In a real-world private-practice series of 100 patients, the author concluded that at least two sessions were needed to reach the aesthetic goal. Stopping after one because the swelling was unpleasant is a common way to end up with a result that does not look like it lasted.
What the fullness is made of: The trials excluded people with excessive loose skin under the chin, and the label tells prescribers to give careful consideration to anyone with marked laxity or prominent neck bands. If part of what you see is skin rather than fat, removing fat will not fix that part, and it will not become durable by repeating the treatment.
Ageing does not stop after treatment either. Nothing in the three-year data speaks to what happens at year five or year ten, because nobody looked.
Recovery, side effects and what the first month is actually like are covered on the Kybella recovery and Kybella swelling day by day pages.