A Few Drops of Blood on Day Two: How Screening Newborns for Spinal Muscular Atrophy Changed a Once-Fatal Diagnosis

What five years of nationwide screening in Taiwan, and a UK planning study, tell families about early detection, treatment timing, and what is still unknown

Why this question matters

Spinal muscular atrophy is a rare inherited condition that weakens the nerves controlling movement and breathing, and it was once the most common genetic cause of death in babies. Three medicines can now change its course — but they work far better when given before a baby shows any symptoms. That single fact is why many countries have begun testing every newborn's heel-prick blood spot for it.

The problem families used to face

For decades, the story followed a painful pattern. A baby seemed healthy at birth. Somewhere between a few weeks and a few months of age, parents noticed the baby was unusually floppy, or was not kicking, or was struggling to feed. Weeks or months of appointments and tests followed. By the time a diagnosis arrived, the nerve cells that carry signals from the spinal cord to the muscles had already been lost — and they do not grow back. Careful studies that followed affected infants without treatment recorded a steady decline, with most children dying or needing a breathing machine before their second birthday [1]. Spinal muscular atrophy was, for a long time, the leading genetic cause of infant death worldwide [2].

The biology behind it is unusually tidy for a genetic disease, which turns out to matter. Almost all cases — about 95% — are caused by the loss of a single gene called SMN1 [3]. Everyone also carries a near-identical backup gene, SMN2, but a tiny difference in its instructions means it produces only a small fraction of the working protein the body needs [4]. People carry different numbers of copies of this backup gene, and in general, more copies means milder disease — though the relationship is not exact [5]. Counting copies of SMN1 and SMN2 is something a laboratory can do from a dried spot of blood, which is what makes newborn screening possible at all.

Since 2016, three medicines have been approved that increase the amount of working protein: one given by repeated spinal injections, one an oral liquid taken daily, and one a single-dose gene therapy [6]. Reviews of the evidence have consistently found that children treated before symptoms appear do considerably better than those treated afterwards [7]. Expert groups accordingly recommend treating screen-identified babies with two or three copies of the backup gene straight away [8]. But by 2023, only 31 countries were screening newborns for the condition, covering roughly 7% of the world's babies [9].

Two studies, and what each one asked

The first study reports five years of a nationwide newborn screening programme in Taiwan, run by Kaohsiung Medical University Hospital together with two of the country's three screening laboratories, with follow-up care at three medical centres [10]. Blood spots were collected from babies' heels at about 48 hours of age. Researchers followed every child identified, some for nearly six years, and recorded what treatment they received, when, and what they could physically do afterwards.

The second is a different kind of study altogether. Before the United Kingdom introduces screening nationally, the UK National Screening Committee commissioned researchers at the University of Sheffield to build a computer model — essentially a careful, transparent set of calculations — estimating how many children would benefit, by how much, and at what cost [11]. Models like this cannot tell you what will happen to any individual child. What they can do, and what this one does unusually honestly, is show which assumptions the answer depends on most, so that the right things get measured once screening actually starts.

What the Taiwanese programme found

Between September 2017 and August 2022, 446,966 babies were tested. Twenty-three were confirmed to have the condition — roughly one in every 19,433 births [10].

The screening itself worked quickly. Babies were flagged at a median age of nine days, meaning half were identified even sooner, and most were seen at a specialist hospital by about ten days of age. Confirming the diagnosis and counting the backup-gene copies took until a median of 21 days.

That timing needs to be held against how fast the most severe form moves. Every baby in the study with only two copies of the backup gene had developed symptoms by 62 days of age, and most by about a month. In other words, for these babies, the disease was already beginning at roughly the moment the confirmatory test result came back.

Here the study becomes a lesson about health systems rather than biology. Taiwan had the screening test years before its national insurance would pay for the medicines: the first was covered in 2020, the second in April 2023, the third in August 2024. So most of these children waited. On average, treatment began at twelve months of age — not because doctors judged that best, but because that was when treatment became available.

The results reflect that delay. Of the 19 children who received treatment, 12 could walk independently, one walked with support, three could sit, one could not sit, and two babies with the most severe form died of sudden breathing failure. Compared with a European group where three-quarters of children were treated before symptoms appeared, a similar proportion of Taiwanese children eventually walked — but far fewer reached that milestone by 18 months. The two children in the Taiwanese group who were treated before symptoms started both walked independently before 18 months, with otherwise ordinary development.

Two other findings deserve mention. Four children had four copies of the backup gene; all four were walking normally, whether or not they had been treated. And one baby's screening test came back normal, yet the child developed symptoms at one month — because that child had a rarer genetic change that this type of test cannot detect. About 5% of cases fall into that category, which is why a floppy baby still deserves investigation even after a normal screen.

What the UK model projects

Using a typical year of 600,000 UK births, the model estimates about 73 babies born with the condition annually [11]. Without screening, almost all of them would be found only after symptoms appeared. With screening, about 69 would be found beforehand.

Over the first three years of life, the model projects that screening would spare two children from needing permanent breathing support, prevent about three early deaths, and mean that roughly 30 children who would otherwise never progress beyond sitting would go further — with about 37 more children ending up with few or no significant physical limitations.

The researchers are careful to name a downside as well. Screening would also identify around three babies a year with five copies of the backup gene — children who may never develop symptoms, or not until adulthood. Telling a family their healthy newborn carries a serious diagnosis they may never experience is a real harm, not a hypothetical one, and the researchers say so directly.

On cost, the picture is genuinely mixed. Screening turned out to cost less overall than not screening, while producing better health — because finding children early avoids expensive complications. But the medicines themselves are extraordinarily costly, and their true prices are confidential commercial secrets, so the researchers had to use published list prices and then test what happened with discounts. Compared against giving no medicines at all, the whole package remains far above what the health service normally considers affordable.

What this means for families

If your baby has a positive screening result, three things are worth knowing. It is not yet a diagnosis — a confirmatory test is needed, and a small number of positive screens turn out to be false alarms caused by unusual but harmless gene arrangements. It is, however, genuinely urgent, and being seen by a specialist within days rather than weeks matters more here than in almost any other newborn screening result. And the number of backup-gene copies your baby has is the single most useful piece of information for what comes next — which is why that test is done immediately.

If your baby has a normal screening result but seems unusually floppy, is not meeting movement milestones, or has weak reflexes, say so to your doctor. A normal screen makes this condition much less likely but does not rule it out.

And if your child has been treated, it is fair to ask for honest expectations. Treatment is not a cure and does not restore nerve cells already lost. In the Taiwanese group, one baby who was already severely affected at birth still needed a breathing tube and feeding tube despite receiving two different medicines. The researchers' own conclusion was that medication alone is not enough for the most severely affected children, and that ongoing care from a team — breathing, feeding, physiotherapy, and family support — remains essential.

What researchers are working on next

The most important unanswered question is durability: does the benefit of early treatment hold at ten, twenty, or fifty years? Nobody knows yet, because the medicines are too new. The Taiwanese follow-up, at a median of five years, is among the longest available anywhere and still cannot answer it.

The second question is speed. If a baby with two backup-gene copies is treated at two weeks rather than at one month, does that change the outcome? The UK researchers have listed this among the priorities for a national evaluation, alongside how accurate the test is in everyday practice, how acceptable screening is to parents, and what the long-term costs really are [11]. Similar structured evaluations have been used before to answer exactly these questions about other newborn screening tests in the UK [12].

The third is fairness. Access to these medicines remains deeply unequal between countries [13], and Taiwan's experience shows why screening without funded treatment delivers far less than it promises.

In summary

Taiwan's five years of data show that a screening test can work beautifully and still deliver disappointing results if the treatment behind it arrives too late — and that the children treated before symptoms began did strikingly well. The UK model, built as that country prepares to start screening, projects real gains in survival and movement while naming honestly what remains uncertain. Read together, they make a point that is easy to miss: the value of newborn screening for this condition is not created in the laboratory. It is created in the days between the heel prick and the first dose.

References

  1. Finkel RS, McDermott MP, Kaufmann P, Darras BT, Chung WK, Sproule DM, et al. Observational study of spinal muscular atrophy type I and implications for clinical trials. Neurology. 2014;83(9):810–817. doi:10.1212/WNL.0000000000000741 ↩
  2. Kolb SJ, Kissel JT. Spinal muscular atrophy. Neurol Clin. 2015;33(4):831–846. doi:10.1016/j.ncl.2015.07.004 ↩
  3. Lefebvre S, Burglen L, Reboullet S, Clermont O, Burlet P, Viollet L, et al. Identification and characterization of a spinal muscular atrophy-determining gene. Cell. 1995;80(1):155–165. doi:10.1016/0092-8674(95)90460-390460-3) ↩
  4. Lorson CL, Hahnen E, Androphy EJ, Wirth B. A single nucleotide in the SMN gene regulates splicing and is responsible for spinal muscular atrophy. Proc Natl Acad Sci U S A. 1999;96(11):6307–6311. doi:10.1073/pnas.96.11.6307 ↩
  5. Calucho M, Bernal S, Alías L, March F, Venceslá A, Rodríguez-Álvarez FJ, et al. Correlation between SMA type and SMN2 copy number revisited: an analysis of 625 unrelated Spanish patients and a compilation of 2834 reported cases. Neuromuscul Disord. 2018;28(3):208–215. doi:10.1016/j.nmd.2018.01.003 ↩
  6. Mercuri E, Sumner CJ, Muntoni F, Darras BT, Finkel RS. Spinal muscular atrophy. Nat Rev Dis Primers. 2022;8(1):52. doi:10.1038/s41572-022-00380-8 ↩
  7. Cooper K, Nalbant G, Sutton A, Harnan S, Thokala P, Chilcott J, et al. Systematic review of presymptomatic treatment for spinal muscular atrophy. Int J Neonatal Screen. 2024;10(3):56. doi:10.3390/ijns10030056 ↩
  8. Glascock J, Sampson J, Haidet-Phillips A, Connolly A, Darras B, Day J, et al. Treatment algorithm for infants diagnosed with spinal muscular atrophy through newborn screening. J Neuromuscul Dis. 2018;5(2):145–158. doi:10.3233/JND-180304 ↩
  9. Vrščaj E, et al. Newborn screening programs for spinal muscular atrophy worldwide in 2023. J Neuromuscul Dis. 2024;11(6):1180–1189. doi:10.1177/22143602241288095 ↩
  10. Wang C-H, Hsu T-R, Liu M-Y, Wang L-Y, Chou I-J, Lee W-T, et al. Newborn screening facilitates early theranostics and improved spinal muscular atrophy outcome: five-year real-world evidence from Taiwan. Orphanet J Rare Dis. 2025;20(1):197. doi:10.1186/s13023-025-03697-1 ↩
  11. Thokala P, Bessey A, Knowles R, Marshall J, Visintin C, Lawton M, et al. Newborn screening for spinal muscular atrophy in the UK: use of modelling to identify priorities for ongoing evaluation. Int J Neonatal Screen. 2026;12(1):3. doi:10.3390/ijns12010003 ↩
  12. Lombardo S, Seedat F, Elliman D, Marshall J. Policy-making and implementation for newborn bloodspot screening in Europe: a comparison between EURORDIS principles and UK practice. Lancet Reg Health Eur. 2023;33:100714. doi:10.1016/j.lanepe.2023.100714 ↩
  13. Armengol VD, et al. Life-saving treatments for spinal muscular atrophy: global access and availability. Neurol Clin Pract. 2024;14(1):e200224. doi:10.1212/CPJ.0000000000200224 ↩