Saturday, January 03, 2009

Important study by Onslow/Packman group

I am known for being highly critical of the Onslow/Packman group for their outcome trials (especially the pseudo random control trial) and other research conducted to prove Lidcombe treatment efficacy. However, this time I have to praise them for their recently published article: Reilly, Onslow, Packman et al. Predicting Stuttering Onset by the Age of 3 Years: A Prospective, Community Cohort Study. Pediatrics. 2009 Jan;123(1):270-277. See abstract here and below:
Participants included a community-ascertained cohort of 1619 2-year-old Australian children recruited at 8 months of age to study the longitudinal development of early language. The main outcome measure was parental telephone report of stuttering onset, verified by face-to-face expert diagnosis. Preonset continuous measures of the child's temperament approach/withdrawal) and language development were available. Information on a range of predictor measures hypothesized to be associated with stuttering onset was obtained (maternal mental health and education levels, gender, premature birth status, birth weight, birth order, twinning, socioeconomic status, family history of stuttering). 
By 3 years of age, the cumulative incidence of stuttering onset was 8.5%. Onset often occurred suddenly over 1 to 3 days (49.6%) and involved the use of word combinations (97.1%). Children who stuttered were not more shy or withdrawn. Male gender, twin birth status, higher vocabulary scores at 2 years of age, and high maternal education were associated with stuttering onset. The multivariable model, however, had low predictive strength; just 3.7% of the total variation in stuttering onset was accounted for. 
The cumulative incidence of stuttering onset was much higher than reported previously. The hypothesized risk factors for stuttering onset together explained little of the variation in stuttering onset up to 3 years of age. Early onset was not associated with language delay, social and environmental factors, or preonset shyness/withdrawal. Health professionals can reassure parents that onset is not unusual up to 3 years of age and seems to be associated with rapid growth in language development.

I do not have the full article but here is the key message:

1) The incidence is 8.5%. We can translate this into a recovery rate if we assume that 1% of adults stutter. 11.8% of 8.5%  do not recover and become the 1%, which means that 88.2% of kids who stutter recover. There will be some statistical error margin and the 1% is an estimate, but the recovery rate should be between 85 and 90%! This is high compared to other studies, but not too far from others sources which refer to a 70 to 80% recovery, and their study should be very reliable, especially considering the very decent sample size of 1600 and their expertise in identifying stuttering.

2) There is no clear and strong correlation with any hypothesized risk factors. They did find higher incidence for male, and other factors. But we need to be careful. I need to have a look at their statistical analysis, given that many variables were tested for and my generally bad experience in former statistical analysis by the group and the field in general. I would not be surprised if only the factor male is very significant.

Here is the implication in my view:

For 1) It must now be clear to everyone that the recovery rate is very high. No wonder therapists can claim such wonderful successes in treating children that would have recovered anyway. Ironically, the BSA and Leys wrote in support of early intervention by claiming a 90% success in treatment: see here. Well you could well get this without treatment, too! (You can still get out by arguing that only severe cases get to treatment where recovery rates are much lower.) The high rate supports my call for considering the natural recovery rate in outcome trials for kids, and this has not been done properly yet. They are all deluding themselves with their outcome trials. Oh and I forgot to say that the famous Lidcombe "random control" trial has also a recovery rate of 84% which fits nicely with their latest article if you assume no treatment effect for Lidcombe!

For 2) It is no surprise to me at all that there is no correlation, and this fits perfectly with the picture that kids start stuttering due to some internal mess-up, either genetic and/or neurological damage. It might correlate to some stages of development but only because this stage happened at the same time when the brain puts it all together to produce automatic speech.


Anonymous said...

Tom -

Thanks for the article info!

88.2% recover naturally...

WOW -- This better not get out to professionals. This would really mess up "claims" their "treating (treatment)" regimens "cure" and/or "prevent" stuttering in children. What would the Queen say?

Now children -- Try turtle talk...Now try it using an easier relaxed approach with slower movement...perhaps a concellation or prolongation would work...Now say it with me using a Speech Easy device...

Anonymous said...

The article sucks big-time, and it's time to get some new people into stuttering research. The current crop of stuttering researchers are just repeating the same tired old experiments that have been done for the past 30 years, and getting more-or-less the same results. How many times have I heard the phrase, "approximately 80% of all infant stutterers will recover spontaneously"? Onslow et al. are just churning out papers to gain research grants, but the papers are worthless.

How can one distinguish between a stutter and a normal "ga ga" baby talk in a 3 year-old baby? At that age, the speech centres in the brain are not fully developed - Broca's area and Wernicke's area are still not co-ordinated. Is this what passes for cutting-edge research into stuttering? If so, I am disgusted!!!! Onslow is a poor excuse for a scientist.

Tom Weidig said...

I agree that most of their research is as you say about "churning out papers to gain research grants, but the papers are worthless."

However, this is a survey on 1600 kids and I do not think that they can mess this one up. And there is no bias in the sample, most likely.

Also, this piece of research is NOT theory-ladden. They do not try to prove something, but only look at all kids.


Tom Weidig said...

>> How can one distinguish between a stutter and a normal "ga ga" baby talk in a 3 year-old baby?

My niece is 2 years and more, but she does not stutter at all like repeating words or show secondary symptoms. I think you can look at 2-year olds and see whether they stutter or not.

Anonymous said...

I also have a niece. When she was 2 years old, she was saying her first words, and I felt certain that she was stuttering. But I was probably trying to look for a stutter that wasn't there; it was just the normal baby talk that all infants do. After that, I did some research on the subject of speech/language acquisition, and came to the conclusion that 2 years old is just too early to be diagnosing a stutter - simply because the brain's speech centres are not yet co-ordinated at that age. I suspect that a lot of slp's may be diagnosing stutters that are not really there in infants ... and then, lo and behold, the kid grows up to speak fluently - spontaneously cured of a stutter that was never there in the first place.
Tom, this is not to say that stuttering is not neurological/genetic in origin. In my very humble opinion, the genesis of stuttering is an over-active amygdala, which has the effect of inhibiting speech production. In a way, a stutter is a result of *both* nature and nurture. I've never seen a better explanation than the following:


Randzig said...

Regarding your last point #2 do you think something like in this video could help. Seems like it would if we are connecting intense experiences with the cause of stuttering.

Olivier said...


The theory exposed in 'veils of stuttering' seems very obscure. The author seems to have complaisance for mysteries, doesn't he ? He write that a medical model of the stuttering might be a bad idea to understand stuttering...This site seems very obscure to me.

This new study is very important, i think it's an advancement.

Anonymous said...

I have to agree with George on this one. This is nothing new and it is a total waste of time.

Olivier said...

Why do you think it's a waste of time ?

Anonymous said...

To Anonymous (January 5, 5:45),

I wouldn't say the article is completely useless. If they manage to print it on soft absorbant paper, I'm sure I could make good use of it. This is the case with all of Onslow's work.


Tom Weidig said...

The research is more important because they try to correlate onset with several hypothesized risk factors without success.

Ora said...

Hey, Anonymous, Anonymous, and Anonymous - why not identify yourselves?

Doesn't have to be your real name, of course, but why not pick a nickname so you'll at least have a consistent identity for the people who read your comments from time to time on this blog?

Anonymous said...

The whole premise of this study (and studies like it) is that SLPs are able to detect a stutter in a 2-year-old infant. For me, this premise is unreasonable; most babies have disfluencies at that age, and it is not an absolute fact that they are due to a stutter. We just assume that some 2-year-old babies stutter, but how do we know? We can't ask the baby, "Hey Mr. Baby, did you just have a block?". And there are no real indicators of a stutter other than the opinion of an slp or researcher. I once spoke to a prominent stuttering "authority" at a convention; as we were chatting, she mistook a slight hesitation of mine as a block. It was definitely not a block in that instance - I was simply searching for the right word. But if such an experienced person can't distinguish a block from a normal hesitation, how is she going to diagnose stuttering in a little baby?

The trouble is that after many decades of "research", nobody knows which part of the brain is the cause of stuttering. If we don't even know that, how can we say that an infant's brain is already developed enough to stutter? This is pseudoscience, and the result is this Lidcombe program (one of the biggest frauds perpetrated on the stuttering community), which is reputed to have an 80+% success rate but, for all we know, it is curing children that never had a problem to begin with.

I may sound harsh, but these are my opinions. Stuttering research has been set back by many decades by people like Onslow, and that is bad for us as stutterers!


Anonymous said...

100% agreement with the poster George and his observations/comments regarding well as his Aussie group of "researchers". I share his opinion.

Kip said...

Hi Anonymous number one,

News flash: Any SLP worth their salt knows the current statistics about natural recovery. The worth of this knowledge is to allow SLP's to empower parents with exactly this knowledge so the PARENTS, not the SLP, can make an informed decision about whether they want to pursue therapy or not.

Anonymous said...

Kip -

Interesting comment you make.

It's my humble understanding that ASHA does not require much - if any - course/clinical work in the area of Fluency and Fluency Disorders in order to allow professionals to deliver therapy with CWS in the public schools.

Wonder where they get their salt to empower parents...

Kip said...

Anonymous -

I'm afraid I can't speak for the American side of things as I'm an Australian. I know that universities in Australia require their SLP students to demonstrate clinical understanding/competence in all areas of speech pathology, including stuttering, before they are allowed to graduate though.

Malcom said...

Kip -

Do you know if SLP candidates in Australia are required to take specific coursework in fluency disorders, and have supervised clinical experiences in treating fluency disorders, PRIOR to graduation from a University in Australia with a degree in S & L?

Thanks in advance!

Anonymous said...

It is certainly interesting for me to read that post. Thanks for it. I like such themes and anything connected to this matter. I definitely want to read more soon.