The power of n….. disease in populations

Sure statistics are a bore for many, but they’re a powerful tool, especially in an age when computerised records of large populations can be linked and crunched with increasing power. ‘n’ represents the number of participants in your study, and in this regard, size matters. The bigger your n, the more statistical power, and the more confidently you can tease out an answer to your question..

sperm

.When you base your beliefs on your cat’s bad reaction to a vaccination, then n=1.  This experience may have strong effect on your view, but it doesn’t necessarily mean vaccinations are bad for all cats. If a vet sees vaccination reactions in 10 of the 30 dogs she saw in a week, then n=30, and she can be more certain there is a batch problem.

Results in man

When n is pushed into the 10’s or 100’s of thousands, we can get some amazing outcomes.  Fiona Stanley’s team, during the mid-90’s, identified maternal B-vitamin deficiency as the cause of the crippling birth defect, spina bifida. Simple  folic acid supplementation in pregnant woman will save 10’s of thousands from suffering.

A similar study of disease in large numbers of parents debunked the ‘perfect sperm’ thesis: winning the race to the egg reflects genetic perfection. The mother’s ovum is no longer blamed for all genetic defects, and the panel beater’s testicles, exposed to the workplace spray booth, stands rightly to account. A win for the sisterhood over habitual patriarchal bias.

.

Picture 1.

Dietary recommendations of less red meat, more fibre, no bacon, more fish, and less char-grilling, are all outcomes of large population studies. ‘Why does daddy have colon cancer?’, rather than left unanswered, can now be at least partially explained by his double-bacon burger addiction.  We don’t know the exact cause of childhood asthma, but maternal smoking, ceasarian delivery, antibiotics, and lack of hookworm exposure have been incriminated by population surveys.

Statistics fingered Vioxx’s rare but catastrophic side effects,  but not before a global marketing campaign saw millions of pills prescribed and billions earned. Similarly, the risks associated with hormone replacement therapy weren’t fully appreciated until the medical fraternity prescribed enough to get a population sufficiently large to prove the breast cancer link.  Glitazones, a family of drugs for treating type-2 diabetes, were hailed as our saviour until  heart, eye, and liver problems started to tarnish their statistical outcomes. It took the birth of about 20,000 thalidomide victims before reaching the numerical proof required for it’s withdrawal from the market.

It’s not just big pharma that has fallen victim to evidence based medicine. Large studies of vitamin E and selenium, hoping to prove antioxidant benefits for cancer and cardiovascular disease, have fallen flat, despite well documented benefits of fresh vegetable and fruit diets naturally high in these compounds.   On the upside,  continue with your fish oil: it’s protective against coronary heart disease and is slowing the growth rate of your prostate tumour.

Fish oil pill

So where are all the veterinary results?

Well, ahem, sorry, we don’t really collect enough data to tell you about the long term benefits or hazards of your dog’s anti-inflammatory or cat’s diet. There are short-term studies done on smaller, experimental populations, for registration, but as described above, more subtle and rare effects aren’t going to be teased out until n is big enough. Compared to human medicine, where there are large, centralised, well-funded institutions for collecting data (universities, hospitals, governments), pharmacovigilance in the veterinary sector is meager.

Big pharma must love the veterinary game. Previcox, a sister drug to Vioxx, has established itself in the vet marketplace as an NSAID with a low side effect profile. While the causes of cardiovascular disease in dog and man are very different,  could the reported sudden death from shock of a dog treated with Previcox in 2007 be  a similar phenomenon to that observed in man? Unfortunately the investigation and numbers required to bring statistical strength to this assertion will not happen any time soon in the vet context.

vioxx

Is cancer in dogs on the increase? Is my cat’s hyperthyroidism due to the iodine content of her pet food? Sorry, we can’t be sure.  Outside of a handful of reportable diseases, there is little or no funding for investigation of unexplained deaths, cancer clusters, birth defects, or the overall prevalence of diseases in animals over time. Almost all research into dietary effects is funded by pet food manufacturers themselves. University teaching hospitals  conduct retrospective studies, looking at rates of disease and treatment outcomes, but again unpooled data results in smaller cohorts, limited statistical power, and few ground breaking results.

During, what was probably, the biggest mass poisoning of cat and dog, the melamine contamination scandal killed thousands and will cripple magnitudes more with premature kidney failure in the future. Involving some of the most reputable brands of  premium pet food, in a nation that prides itself on it’s level of animal welfare and veterinary care, the US Food and Drug Administration can only confirm 14 cases out of 3,600 reported deaths. For more on this, see The Argument for Dietary Diversity.

In their defense, no government has the cash to invest in medical investigation and centralised recording of companion animal illness or death, unless it has implications for public health in man. Even disease in economically important species, the ones we eat, is falling victim to budget cuts in Australian Departments of Agriculture.

Pet owners and vets can voluntarily report adverse events like reactions to vaccines and drugs, but such a task easily drops off the bottom of the to-do-list in a busy clinic. In defense of the under-resourced local vet – unless a side effect is severe, immediate, and known to be associated with a particular drug, its hard for us to make causal links. Is  rupture of a dog’s cruciate ligament associated with my prescribing a course of  fluoroquinolone antibiotic, as it has been in man? We may never know.x

Tags: , , , , , , , ,

Leave a Reply