An Ounce of Prevention?

Last updated on September 23, 2018

How much impact would SCA prevention make? I have developed a web-based tool to help look at the question.

DNA testing has been available up through SCA3 since 1995. I think this influenced some people’s approaches to family planning—at least, it did mine; I was tested in 1996. If you know the disease is genetic and have proven that you have it with a DNA test before you have offspring, then only if you’re willing to pass on the defect and disease to your offspring would you have offspring with no safeguards. This case is head-scratchingly common today. Many are fine with the 50/50 odds and refuse to have testing in their lives or their minor children.

For centuries prior to 1995, people were passing on the SCA3 defect to their offspring without safeguards or even knowledge available. Apparently, that’s when the prevalence range of 1-5 per 100,000 for all SCAs combined was established.

Let’s first think about the range from 1 to 5

One to five is a gigantic range, perhaps to the point of being meaningless. It’s psychological trickery. The increase from 1 to 5 is a 400% increase; the decrease from 5 to 1 is an 80% decrease; and many points in between. Here’s a table to keep us reminded:

Since ratios are involved, a change from 5 to 1 is not just a small decrease of four nebulous little “points:” it’s a decrease to one-fifth. Here are two examples of the same ratio applied six times each, to put the decreases into a bigger numerical context:

80% decrease: 625 → 125 → 25 → 5 → 1 → 0.2 → 0.04 → 0.008

20% decrease: 9.8 → 7.8 → 6.3 → 5 → 4 → 3.2 → 2.6 → 2.0

Stock day traders and penny stock traders worth some salt will be familiar with numbers like these. If you buy a stock at $5/share and it decreases by 80% to $1/share, it must increase 400% to get back to $5/share. There’s a distinct lack of parallelism that plays tricks with the mind.

I mean, really

One thing I don’t consider here in detail is real numbers vs. integers. There’s a difference in precision between 1.0 per 100,000 and 1 per 100,000. And 1 per 1,000,000,000 might be as good as zero in the U.S.

Zero and Zeno

The space between 0 and 1 is the weirdest of all. In a sense, we bump into what’s known as Zeno’s paradox: we can get infinitely close to zero without reaching it. This is beginning to spell trouble for the impact of SCA prevention:

Zero → any-positive-number = increase of positive infinity

Any-positive-number → zero = 100% decrease

More weirdness

Not all integers are created equal.

Temperature. Let’s say you want to compare 0 and 5 degrees Celsius temperatures. In this case, you probably want to convert first to Kelvin, and then do the comparison.

0 °C = 273.15 K
5 °C = 278.15 K

If you’re concerned with temperature as defined by the activity of atomic and molecular particles, then 0 → 5 °C is an increase only by 1.83%.

Medication. Let’s say you take 100,000 mg of acetaminophen. Let’s say that will kill you—I don’t know what a lethal dose might be. Instead, let’s say you take 10 mg of acetaminophen. The body will likely ignore that small of an amount and flush it out without even a blip. Does that mean you have come within 0.01% of death? Probably not, but this shows how meaningless and absurd it can be to compare numbers.

What’s the question again?

What is the effect on prevalence of a falling disease birth prevalence (i.e., incidence)? How long does change take?

Rough math for this is just within my reach. Coupled with some iterative JavaScript, I produced a web page that shows a few things in action.

Here’s the idea behind one iteration:

One thing to notice is that the size of the population does not matter.

To keep things simple, I have used a constant birth rate and constant death rate, rather than using functions that change over time (in reality, both numbers have been falling). Here’s the page:

http://ataxia.hemiola.com/prevention/

How long for prevalence to drop?

Let’s say the SCA incidence dropped to zero. How long would it take for prevalence to drop? In the following table, the row is the starting prevalence, and the column is the resulting prevalence; the numbers in the table body are in years.

How much drop in 22 years?

It’s now been about 22 years that SCA1-SCA3 tests have been available. If we went for 22 years without a single SCA birth, this would have been the (approx.) impact on prevalence (per 100,000):

5.0 → 3.2
4.0 → 2.6
3.0 → 1.9
2.0 → 1.3
1.0 → 0.6

In other words, a prevalence drop of about 36% was missed out on because of people’s willingness or obliviousness to passing it on.

Conclusions

“Perfect” prevention won’t quickly eradicate SCA from the population; it will take decades. There are no counterintuitive surprises that I see.

It is my guess that people will keep unknowingly having SCA offspring, deep into the future. And if genetic therapies are developed that ameliorate symptoms, that will further assist against prevention.

There are too many societal issues here. My position is as follows:

• If you are born with SCA, you will eventually suffer the consequences.
• If you are born with SCA and know it, then if you knowingly don’t pass it on, you have done the best you can towards prevalence reduction.

Pondering more cases than that gets too complicated.