Etiology Rationale

Congenital hearing loss can be broken down into genetic/hereditary causes and pregnancy related causes. Genetic/hereditary causes can further be broken down into syndromatic (issues with more than just hearing loss) and non-syndromatic (just hearing loss).

Gallaudet Research Institute Study

One of the biggest challenges I found was that while there are 8,107 students listed with genetic causes of hearing loss, only 3,143 reported the specific cause, and 2,237 of them listed "Other" as the cause. My guess is that there are many children who have other deaf people in their family who assume that the deafness is genetic, but who have never been tested. So my guess is that the 2,237 listed as "Other" are really mostly "Unknown". This distinction is very important.

Another related issue is that there are too many of some categories, but not enough of others. For example, 4.7% of students did not list any information on the cause of hearing loss. Of those that did, 53.7% listed "Cause cannot be determined/DNA" (DNA presumably does not mean 'did not answer', as that sounds the same as students not reporting any information). Of the 22.7% that listed 'Genetic/Hereditary/Familial' causes, 61.2% did not list a specific cause, and of those that did list a specific cause, 71.2% listed 'Other'. There were no options for a couple of common syndromes.

One inherent flaw in using this data for determining etiology is that it is old information. Specifically, the median age of the children appears to be around 12 years old for this study that appears to have been conducted between 2004 and 2005. That means that the information is about children who were born around 1992. At that point, genetic testing wasn't available, which likely accounts for the abnormally small percentage of students listing known non-syndromatic genetic causes of hearing loss. And, newborn screening was not as common, making it more difficult to determine many congenital causes.

However, despite its flaws for our purpose, it provides some very useful information.

Overall Challenges

Also, many sources refer to a syndrome or other issue as accounting for a certain percentage of "all childhood deafness", which has to be interpreted. Do they mean just congential causes, or do they include acquired causes? If they include acquired causes, which ones do they include? Ones that start before age 18?

Another issue is that there are some rare syndromes that have a high mortality rate, so numbers appearing in surveys of school-age children might underrepresent these syndromes.

Speaking of syndromes, there's the problem that some of them do not cause hearing loss until after the person has passed early childhood. Although the syndrome is congenital, the hearing loss won't be known at or near birth, so we have to decide if those should be included.

Yet another issue is the degree of hearing loss. First, it may be that a lot of children with mild hearing loss do not make it into the studies (for example, if they are mainstreamed into the normal school system). And, some causes of hearing loss are reported to be the cause of hearing loss in a certain percentage of "profoundly deaf" children. Figuring out what percentage of children are profoundly deaf is challenging. Finally, there is no exact cutoff for what constitutes hearing loss.

Then, there's the whole issue of adding up conflicting numbers. Let's say that you know that exactly 15% of congenital hearing loss is caused by syndromes, yet after adding up the numbers for all syndromes, it comes out to 18%. What should be done there? Perhaps in this case syndromatic hearing loss accounts for a higher percentage that was thought. Or maybe the percentages for one or more individual syndromes were overstated. Somehow, this needs to be worked out to get the numbers to add up to 100%.

And, the big old "unknown" category is biased. For example, while about 50% of the time the cause of hearing loss is not known, that 50% isn't evenly spread among all causes. For example, if hearing loss was caused by drug abuse by the mother during pregnancy, the mother likely wouldn't want to admit that. And some syndromes are easily identified in most cases, so not much of the "unknown" category would be due to these syndromes.

Finally, if things weren't confusing enough already, what happens if there are multiple causes of hearing loss? This should be a rare event (if 1 in 1,000 children are born with hearing loss, that would suggest that only about 0.1% of children with hearing loss have multiple causes). But it is something that needs to be taken into consideration.

The Starting Point

First Breakdown: Genetic v. Environmental

Next: Syndromatic v. Non-Syndromatic

Next: Dominant, Recessive, X-Linked, Mitochondrial

I'm going with 78% recessive, 20% dominant, 2% X-Linked or mitochondrial. 75% to 80% seems to be the best range for recessive, 1-2% for X-Linked and mitochondrial, leaving about 20% as dominant.

Next: Recessive GJB2 vs. Recessive Other

Breaking Down Environmental Causes

For this, we really need to rely on the Gallaudet study, as there is precious little other information that is easily accessible. While there are plenty of experts on genetics to do studies and provide information, there don't seem to be many experts that study long-term consequences of issues during pregnancy.

Gallaudet breaks down environmental factors into rubella, CMV, drug/alcohol abuse, Rh incompatibility, consequence of prematurity, trauma at birth, and other pregnancy complications (which could include Toxoplasmosis...). It's that "other" category (accounting for 31% of environmental causes in the study) that makes things difficult.

For toxoplasmosis, an estimated 400-4,000 cases occur per year (in the US), of which about 10-15% have hearing loss. Given about 12,000 newborns with hearing loss in the US, that's 0.3% to 5% of congenital causes. Estimating near the low end gives us 0.5%.

For Herpes Simplex, it is estimated that about 1 in 3,000 to 20,000 live newborns have HSV, with 15-30% having HSV-1, and about 50% of those with HSV-1 having bilateral sensorineural hearing loss. Erring on the safe side, that's about 1 in 100,000 live births compared with about 1 in 500 having hearing loss, or about 0.5%.

Syphilis affected about 1 in 7,500 newborns in the United States in 2000. This is from the CDC, who got the information from the year 2000 natality data. However, it appears that the data the CDC has is not included in the publicly accessible data, so I could neither confirm it nor check for more recent data. It sounds like syphilis is on the decline, but without more recent data, I have to go with the 2000 data. One source shows up to 25%-38% having hearing loss, but with the hearing loss appearing before 10 years old in only 37% of them. So let's go with 25% having hearing loss, and make a guess that 10% have hearing loss at or around birth. That's 2.5% or 1 in 40, or about 1 in 300,000 newborns. That would make congenital syphilis accountable for about 0.16% of congenital hearing loss (and declining).

Rubella may no longer be a significant cause of hearing loss in developed countries. It appears that in 1970, it may have accounted for about 0.5% of congenital hearing loss, but in 1996 there were only 2 cases of congenital rubella in the United States, and only 5 cases between the start of 2001 and the end of 2004.

CMV (cytomegalovirus) is very common, affecting about 1% of all newborns (about 30,000 to 40,000 newborns in the U.S. each year). One study showed about 5% of cases resulting in hearing loss at birth (with another 10% having early onset hearing loss by 6 years old), another source claims about 7%. Erring on the safe side with 30,000 newborns and 5% having hearing loss, that's 1,500 newborns each year with hearing loss due to CMV. That's about 12.5% of all congenital causes!

Hyperbilirubinemia is next on the list, with kernicterus apparently being the main known problem. Jaundice (the yellowing that is seen with hyperbilirubinemia) is extremely common in infants. If it affects the neurological system, it is considered kernicterus. The CDC states a study showing the prevalence of kernicterus as 7.5 per 100,000 live births. Another source shows 20% as having "severe deafness." That works out to be the cause of 2.5% of congenital hearing loss. Given the likelihood of more than 20% of those with kernicterus having any degree of hearing loss, and the liklihood that many infants with hyperbilirubinemia (but without kernicterus) have hearing loss, 2.5% seems a very conservative number.

Bacterial Meningitis is yet another cause of congenital hearing loss. Here, we see it occurring in 0.3:1000 live births, with about 25% having hearing loss by age 5 and 30% at any age. One source suggests that a large percentage of the hearing loss is present at or shortly following birth. Assuming half of those that end up with hearing loss have it at birth, that would be 3.75 cases of hearing loss caused by bacterial meningitis per 100,000 live births. That's about 1.25% of all congenital hearing loss.

Hypoxia (inadequate oxygen supply), anoxia (no oxygen supply), and asphyxia (inability to breathe normally) may be a relatively common cause of congenital hearing loss, but data is hard to come by. One study showed 4.5 per 1000 births in California having asphixia, although the numbers declined to about 1.3 per 1000 in the year 2000 (down from 14.8 in 1991). Two sources show about 20% as having hearing loss, another shows just 3%. Going with 1.3 per 1000 and 10%, that leads us to 4.3%, a surprisingly high number.

Rh Incompatibility is one that I'm not including, since it appears that Rh incompatibility often results in kernicterus, which causes the hearing loss, and which we already included.

For Drug/Alcohol Abuse, including Fetal Alcohol Syndrome, there isn't much data as to what percentage of affected infants have hearing loss. Given the lack of data, I'm going to start with the data (0.7%) from the Gallaudet study (which I believe may be on the low side, especially since many parents may not have realized why the deafness occurred, or may not want to let the children or others know). Factoring in the prevalence of Fetal Alcohol Syndrome (perhaps 1:1000), that would mean that about 2% of infants with FAS have hearing loss, which seems low. So I'm going to double the Gallaudet number, with a 1.4% figure, and add in 0.6% for other drug-related hearing loss. Given underreporting, I think these numbers are safe.

For toxemia (pre-eclampsia), it doesn't seem to be a common cause for hearing loss per a study linked to by the NIH.

Maternal Diabetes does not seem to be a cause of hearing loss, even though it is listed in several sources as a cause of hearing loss. Although there is a "maternally inherited diabetes and deafness" (MIDD), it appears to be extremely rare (or an "unknown cause"), and it appears as though the hearing loss may not normally be present at birth.

That's all the environmental congenital causes I found, and they added up to less than the 40% total for environmental congenital causes. So I just added an 'Other' category. Now we have:

Breaking Down Syndromatic Causes

Down Syndrome - About 1 in 800 children are born with Down Syndrome, and studies show between 60-80% having hearing loss. It's not clear how much is present at birth, but 25% might be a safe guess. That puts it at about 10.5% of congenital hearing loss. The Gallaudet study shows 8.7%, which seems safer. True, that 8.7% is including hearing loss that occurred after birth, but the 8.7% figure seems very low compared to the 25% that the statistics would otherwise suggest, so I was going to go with 8.7%.

But, given the discrepancy after totalling the numbers, I'm going to reduce this to 6%, to account for less than 25% of hearing loss present at birth.

Usher Syndrome appears in about 4 per 100,000 births, with hearing loss generally present at birth. Some sources show Usher as accounting for 5% to 10% of congenital hearing loss -- but that seems to be completely inaccurate, given that it seems that there are no more than 3-4 cases per 100,000 births. That would account for about 1.3% of all congenital hearing loss (which the Gallaudet study confirms).

Pendred Syndrome is tricky, since the prevalence is unknown (one source shows it as being between 1:10,000 and 1:100,000 live births). Sources seem to agree that it accounts for roughly 5-6% of congenital hearing loss. That would mean that it would account for about 1:6,500 live births. The Gallaudet study doesn't have information on Pendred. It seems that about 80-85% have hearing loss at birth, so I wanted to go with Pendred accounting for about 5% of congenital hearing loss. But, given the discrepancy after totalling the syndromatic causes, I'm going to assume 1:10,000 is correct, and go with 2.8%.

Waardenburg occurs in about 1:20,000 to 1:50,000 live births, with about 30% having hearing loss. At 1:20,000, that would be 0.5%. Yet Gallaudet shows about 2%. Digging further, a few sources show 1:4,000 live births, with a book by suggesting a higher incidence. At 1:4,000 and 30% with hearing loss at birth, that would be 2.5% of all congenital hearing loss, which seems plausible.

CHARGE Syndrome affects roughly about 1 in 10,000 newborns, with about 60% having hearing loss. That translates to about 2% of all congenital hearing loss. Gallaudet shows about 2.5%, which would correlate nicely with the 1:8,500 number seen in some sources. Unless the incidence of CHARGE syndrome is decreasing, 2.5% seems appropriate.

Treacher Collins Syndrome affects about 1 in 25,000 to 50,000 newborns, with about 50% having hearing loss. That's 0.33% to 0.66% of hearing loss, yet Gallaudet shows it as about 1.4%. One source shows as many as 1 in 10,000 newborns having Treacher Collins, which would be 1.6%. 1.4% seems a bit on the high side, so I'm going to go with 1.0%, somewhere in the middle.

Branchio-Oto-Renal Syndrome affects about 1 in 40,000 newborns, with about 75% having hearing loss. That's about 0.6% of congenital hearing loss.

Crouzon Syndrome affects about 1:60,000, with about 30% to 55% with hearing loss, accounting for about 0.15% of congenital hearing loss.

Jervell & Lange Syndrome affects about .16 to .6 out of 100,000 newborns, with hearing loss in all (or most?). That's about 0.1% to 0.2%. I'll go with 0.1%. Stickler Syndrome affects between 1:7,500 and 1:10,000 live births, with 15% to 40% having hearing loss that may be progressive. So let's go with 1:7,500 and 15%, to come up with .6%. Goldenhar Syndrome affects between about 2 to 14 out of every 100,000 live births. Going with 5 per 100,000, and 1 in 3 having hearing loss at birth, that would account for about 0.5% of all congenital hearing loss. That's the syndromatic causes I found, but they add up to 22.9%, but we have syndromatic hearing loss accounting for 18% of congenital hearing loss. So either we've overestimated how common some syndromatic causes are, or we've underestimated how much syndromatic causes contribute to congenital hearing loss. I'm going to reduce Pendred from 5% to 2.8%, bringing 22.4% down to 20.2%. Then, I'm going to reduce Down Syndrome from 8.7% to 5.5%, as I may have overestimated how common the hearing loss is at birth (rather than occurring later), bringing the total down to 17.5%. I'll add an "other" category for .5%.

Finally, sort the table!

Other causes to consider...