Fact Check: New Finnish "Study" Does Not Prove "Trans Youth Care Leads To Worse Outcomes"

The latest Kaltiala study is filled with fatal methodological flaws, and does not support claims made about trans youth care lacking efficacy.

Puheenaihe

A new study circulating on social media falsely claims that gender-affirming care for youth and adults under 23 does not improve mental health outcomes—and may have even worsened them. The study, published in the low-impact journal Acta Paediatrica, comes from Riittakerttu Kaltiala, a Finnish psychiatrist with extensive ties to at least one anti-LGBTQ+ hate group and a history of anti-trans political advocacy. The study uses Finnish registries and psychiatric visit data to support these conclusions. However, a closer examination reveals the study to be fatally flawed: it does not actually measure what it claims to measure, its headline finding is a massive artifact of surveillance bias, the clinics practices were abusive, and it operates within a system where its findings were essentially baked in from the start—regardless of what the actual mental health impacts of gender-affirming care might be.

The study looks at 2,083 people who walked into one of Finland's two gender identity clinics between 1996 and 2019, before turning 23. These patients were split into those who ultimately received gender-affirming care and those who did not. The study sets its point of comparison—what it calls the "index date"—at the date of a patient's first appointment at the clinic, not the date any treatment began. It then looks at whether each person had any visit to a specialist psychiatrist before and after that date, which it labels "specialist-level psychiatric treatment." Its headline finding: among trans girls and women who received gender-affirming care, "psychiatric treatment" went from 9.8% before to a staggering 60.7% after. The numbers sound devastating—except the study commits several fatal errors. Its central finding is not supported by its own data. The outcome variable conflates merely visiting a psychiatrist with actual psychiatric deterioration, something the study had the data to disentangle but chose not to. Most damningly, Finland's own medical journal documented that trans people were monitored every 3-6 months "in close cooperation with the adolescent psychiatry unit"—visits that register in the very same database the study uses as its outcome measure. While the study claims to exclude gender identity team appointments, it cannot exclude what those appointments produce—any downstream referral to a separate specialist for evaluation, even one that clears the patient entirely, registers in the study's outcome measure.

A closer examination shows that the study’s conclusions are entirely unsupported by its own data. This is part of a pattern from Kaltiala, who published a similarly flawed study in 2024 that used proxy variables and buried its own positive findings to claim gender-affirming care does not reduce suicide—when its data actually showed the opposite. That the current study landed in Acta Paediatrica, a journal with an impact factor of just 2.1—far below top-tier pediatrics journals like JAMA Pediatrics or Pediatrics—suggests it may not have survived scrutiny at more rigorous outlets. This fact check will address the study’s most severe methodological shortcomings, to counter the misleading narrative already spreading across social media and anti-trans policy circles.

The Study Measures Psychiatric Appointments That Are Virtually Guaranteed, A Case Of Massive “Surveillance Bias”

The study measured a single binary variable to determine if gender-affirming care impacts transgender youth negatively: did this person have any contact with specialist-level psychiatric services, yes or no? If a person had even a single contact with a specialist-level psychiatric service, they were recorded as needing psychiatric care. Notably, the study calls this variable “need for specialist-level psychiatric treatment”—but the registry it draws from, the Care Register for Health Care, records visits, not treatments, and the study itself notes this limitation:” A limitation of this study is that more detailed information on patients' reasons for using the psychiatric services could not be analysed.”

What this means in practice is that a single specialist referral generated by a routine monitoring appointment registers identically in this dataset to a psychiatric hospitalization. An external psychiatric evaluation prompted by a flag at a quarterly check-in—even one that clears the patient entirely—counts the same as an emergency crisis intervention. The study excludes the gender identity team's own appointments, but the web of downstream referrals those appointments generate is fully counted.

This is because Finland's gender identity services for minors are not housed in a general medical setting—they are located within the departments of adolescent psychiatry at both Tampere and Helsinki University Hospitals. The assessments are conducted by psychiatric teams: an adolescent psychiatrist, a psychiatric nurse, a social worker, and a psychologist. In other words, the gender care pathway in Finland IS a psychiatric pathway and deeply enmeshed with the psychiatric system. Patients were monitored every 3-6 months during hormone treatment, and at each check-in, clinicians actively screened for psychiatric concerns. Any flag—however minor—could generate a referral to separate specialist psychiatric services outside the gender identity team. Those referrals are not excluded from the study's outcome variable, even if the specialist evaluation found no disorder at all.

Notably, most transgender youth in the program will be subject to psychiatric appointments, both before, during, and after the “index date.” Anything found during these appointments will require continued care and evaluation. This is a textbook case of surveillance bias—a well-known problem in epidemiology where a group under clinical observation appears sicker than a comparison group simply because they are being watched more closely. Trans youth in Finland’s gender care system are, by design, in constant contact with psychiatric services.

To understand why this matters, consider an analogy. Imagine a study that measured whether cancer patients had "contact with oncology services" before and after being referred to a cancer center, and then compared them to randomly selected healthy people. You would obviously find that cancer patients have more oncology contact than healthy people, both before and after referral. You would find that cancer patients who received treatment (chemotherapy, surgery) had increased oncology contact after treatment—because treatment requires monitoring, follow-up scans, medication adjustments, and post-surgical care. Would you conclude from this that cancer treatment "doesn't work" or "makes cancer worse"? Of course not. But that is exactly the logic this study applies to gender-affirming care.

To put the full picture together: a trans teenager walks into a Finnish gender clinic for the first time. That date becomes the study's "index date." The study waits two years before it starts counting psychiatric visits—but as we have established, the diagnostic process alone took one and a half to two years, and often longer. Before they can even begin treatment, any co-occurring psychiatric conditions must be stabilized, adding months or years. After diagnosis, additional psychotherapy check-ins are required before hormones can start. For many patients, the two-year mark arrives while they are still in the middle of the assessment—or have only just begun hormones. And once hormones do start, the Finnish Medical Journal confirmed that patients are monitored every three to six months in "close cooperation with the adolescent psychiatry unit." When the study checks whether a patient had "any psychiatric contact" after the two-year mark, the answer is virtually guaranteed to be yes. Quarterly monitoring check-ins with a team known for aggressive screening would inevitably generate downstream referrals to external specialists—referrals that are not excluded from the study's outcome variable. Even a referral that results in the patient being cleared of any psychiatric issue counts identically to a psychiatric hospitalization in this binary measure.

Immortal Time Bias: The “Before And After” Is Not “Before And After Treatment”

Another major flaw is the study’s choice of “index date”—the point in time around which the entire before-and-after comparison is built. The study sets this date at a patient’s first appointment at the gender identity clinic. This is not when treatment began. In Finland’s highly restrictive system, the assessment process and psychiatric care can and does stretch over multiple years. Patient accounts from Kaltiala’s own clinic describe years-long diagnostic processes before any treatment decision is made.

The study’s own methods section reveals that the authors had the tools to fix this. It states that hormonal treatment was identified through the Social Insurance Institution’s records of prescription medication purchases. If the authors used these records to determine who received hormones—which is how they defined their treated and untreated groups—they necessarily had access to when those prescriptions were first filled. They could have anchored the before-and-after comparison to the actual start of treatment rather than the clinic door. They chose not to.

This would be a serious problem on its own. But it becomes far worse when combined with a well-known statistical error called immortal time bias. In simple terms, immortal time bias occurs when a study classifies someone as “treated” for a period of time during which they had not yet received treatment. The study categorizes patients as GR+ or GR- based on whether they ever received gender-affirming care during the follow-up period—but it does not account for when that care actually started. A patient whose first clinic appointment was in 2012 but who did not begin hormones until 2016 is coded as GR+ for the entire period, including the four years before any treatment occurred.

During those four years, this patient is navigating Finland’s lengthy assessment, undergoing required psychotherapy, and being monitored—all of which can easily generate specialist psychiatric visits that register in the same database the study uses as its outcome measure. Those pre-treatment visits get attributed to the “treated” group, even though they occurred before treatment. And because the study uses a binary outcome—any visit, yes or no—a single required psychotherapy appointment during those pre-treatment years is enough to flip the outcome to “yes” and be counted as evidence that gender-affirming care led to “psychiatric treatment needs.” The binary measure and the immortal time bias compound each other: the longer a patient waits for treatment inside the system, the more guaranteed it becomes that they will accumulate at least one psychiatric visit that gets misattributed to the effects of care they had not yet received.

Kaltiala’s Abusive Clinic May Have Directly Led To Increased Psychiatric Visits

There is another important dimension to the psychiatric contact increase that goes beyond surveillance bias: the likelihood that Kaltiala’s clinics and the assessment process itself are generating psychological distress that would require psychiatric care.

Kaltiala is the chief psychiatrist in the department of adolescent psychiatry at Tampere University Hospital and has run one of Finland’s only two gender clinics for minors since its creation in 2011. Despite having little to no prior research portfolio in gender medicine when she was tasked with establishing the clinic, she has overseen virtually every aspect of adolescent gender care at Tampere for the past 15 years.

Multiple former patients have described the experience at her clinic as abusive. Patient accounts collected by Finnish LGBTQ+ publication Kehrääjä and reported by this publication paint a disturbing picture. A 13-year-old trans girl, identified as “Milla,” was asked by a male doctor in a room with three unfamiliar adults to describe what she thought about her naked body while looking in a mirror. When she began to cry, the staff was reportedly confused by her reaction. The same patient was pressured to discuss her erections and urination with strangers. Another patient was asked to “describe to me how you masturbate”—a question rooted in the discredited theory that transgender identity is a sexual fetish. Multiple families reported that their children were systematically called by birth names and referred to by their sex assigned at birth throughout the process, even after legal name changes and even when the child had been living in their gender for years. One mother told Kehrääjä that her daughter was treated as a “cisgender boy” by the clinic from start to finish.

A trans boy identified as “Mikko” was denied a diagnosis after years in the system because his hobbies were deemed “too girlish” and that he would be “too short” for a man and warned at every appointment that the process “could be terminated at any time.” One family described the entire approach as a search for reasons the child could not possibly be transgender: “It feels like the whole process is about searching for various reasons why the child is transgender. The option that transgender identity could be an innate, unchangeable trait is considered only as a last resort.” When one family obtained a legal diagnosis and prescription for puberty blockers from a private doctor, the clinic reported them to child protective services and filed a criminal complaint. No legal consequences resulted, but the clinic terminated the patient’s care. These are the clinics that generated the data in this study—and the psychological harm caused by these practices would itself register as exactly the kind of specialist psychiatric contact the study measures as its outcome variable.

The Bottom Line

Though this study has made waves on social media, there is a reason it did not land in a premier scientific journal. Kaltiala is not a detached observer of transgender care—she is one of the most prominent anti-trans activists in the medical world. She has received honoraria from an organization designated as an anti-LGBTQ+ hate group by the Southern Poverty Law Center, served on the advisory board of the Cass Review, testified as a primary witness in favor of Ron DeSantis’ ban on transgender care in Florida, facilitated secret meetings between DeSantis appointees and British health officials, and presides over clinics where patients have reported abusive practices including invasive sexual questioning of minors. This is the second time in two years that Kaltiala’s team has published a study engineered to reach a predetermined conclusion and promoted through the same pipeline—from SEGM to Ben Ryan to Matthew Yglesias—within a day of publication.

The data in this study does not say what the authors claim it says. They used a proxy variable that measures clinic visits rather than mental health. They relied on data saturated with surveillance bias from a system that requires the very psychiatric contacts it then counts as evidence of deterioration. They anchored their timeline to the clinic door rather than to treatment, despite having the data to do otherwise. And they did all of this in a system where the result was structurally guaranteed before a single data point was collected. The study’s fate will not be determined by its scientific merit—which is minimal—but by whether the political infrastructure that surrounds it succeeds in laundering it into policy.

Update 4/7/26:

Dr. Kaltiala has now responded to our criticisms of her study through Gender Clinic News, an anti-trans blog whose author describes transgender healthcare as "gender ideology" and refers to advocates of gender-affirming care as "gender ideologues.” Her response is notable less for what it addresses than for what it avoids entirely. Of the major methodological flaws identified in this fact check—surveillance bias, the misanchored index date, immortal time bias, the binary outcome measure, and the abusive clinic practices—she engages with almost none of them.

She does not address surveillance bias at all. She does not address the fact that her study pegged its before-and-after comparison to the date (+2 years) a patient first walked into the clinic—not to when treatment actually began, and how this creates an immortal time bias. She does not explain why, despite having access to Social Insurance Institution records showing when each patient first filled a hormone prescription, she chose not to anchor the comparison to treatment. She does not address the fact that for many patients, the two-year gap after the index date expires while they are still mid-assessment. She does not address how the binary outcome measure—any visit, yes or no—means a single downstream referral from a routine monitoring check-in is enough to be counted as evidence of “psychiatric treatment needs.” She does not address the abusive clinic practices documented by former patients at her own facility, or the patient accounts collected by Kehrääjä describing invasive sexual prodding of minors, systematic deadnaming, reporting of trans youth to CPS, and abuse. These are fundamental design choices and contextual realities that predetermined the study’s outcome, and her response is silent on all of them.

To the one substantive rebuttal she does make: she claims that Finland’s specialist psychiatric services are reserved for those with “severe disorders” and that patients would not reach specialist-level care without genuine need. But this misses the point of the surveillance bias critique entirely. The study claims to exclude appointments with the gender identity assessment team itself—but it cannot exclude what those appointments produce. The Finnish Medical Journal documented that hormone therapy patients were monitored every 3-6 months “in close cooperation with the adolescent psychiatry unit.” At each of these regular check-ins, clinicians are actively screening for psychiatric concerns. If anything is flagged, and notably this team had a hair trigger for flagging concerns, the patient is referred to separate specialist psychiatric services for evaluation and treatment. Those downstream referrals are not excluded from the study's outcome variable. If a referral is made for any concern at all, the patient may continue to require visits long after, even for conditions entirely unrelated to gender dysphoria, such as ADHD or unrelated trauma. Damningly, even if the specialist determines there are no issues and clears the patient entirely, the visit is still logged, the binary still flips to "yes," and the study counts it as evidence of "psychiatric treatment needs" caused by gender-affirming care.

This is textbook surveillance bias: when you screen a population every three months for psychiatric issues, you will find and refer more people than in a population that is never screened. The general population controls in this study had no equivalent system watching them. The “dramatic increase” is not evidence of deterioration—it is what happens when you take a group selected for having almost no prior psychiatric contact and place them inside a system that screens them every three months and refers them onward routinely.

Dr. Kaltiala’s own study concedes the point that defeats her rebuttal. Her study states: “A limitation of this study is that more detailed information on patients’ reasons for using the psychiatric services could not be analysed.” Her response to Gender Clinic News repeats this: the data “included diagnoses and in-patient periods but not treatment nor information about a patient’s problems.” She is acknowledging, in her own words, that she cannot distinguish a visit driven by genuine psychiatric crisis from one generated by the monitoring system built into her own clinics. She cannot tell whether a specialist referral reflects a severe disorder or a routine flag stemming from a quarterly check-in, or even a specialist assessment that evaluated and cleared the patient of any psychiatric issues entirely—all register identically in the database, and all count the same in her binary measure. She cannot tell whether a patient’s psychiatric contact was caused by gender-affirming care, by the assessment process, by the abusive practices documented at her clinic, or by a pre-existing condition that would have been caught in any population subjected to this level of screening. And yet she draws a causal conclusion anyway. That is not a limitation. It is the entire reason her study is fatally flawed and cannot draw the conclusions she and others are drawing from it.