Thursday, November 21, 2013

Evaluating Nutritional Research: The Problems Meg Smart DVM, PhD

Research Funding

A recent article “Science under Siege” although about pharmaceutical research(Discover Magazine Oct2007)) reflects what is happening in the pet food industry . Private funding to academic institutions by big pharmaceutical companies is allowing science to become a powerful tool in their fight against regulation. Research in small animal nutrition has been traditionally underfunded or more accurately seldom funded by independent granting agencies. This has left the field wide open for the pet food industry to control and direct the research done in an academic institution, and within their own facilities. Research into pet foods is seldom at “arm’s length”.


Understanding and Interpreting Research and Development

Research into the nutritional requirements of companion animals is not as straight forward as it is in
livestock nutrition. In food animal trials the end point of the experiments are set with defined
production parameters that are , measured, statistically compared, and conclusions drawn from a
relatively uniform group of animals. Although all animal trials must have prior approval by affiliated
Animal Care Committees to ensure humane treatment, livestock trials often, end with the humane
slaughter of the animals and tissues harvested for detailed analysis. In the past and to a limited degree
now, the animals were surgically altered fed poisons directed to damage a specific organ in order to
evaluate the response of that animal to nutritional intervention. Although the results are still used to
justify the formulation of some Veterinary Therapeutic Diets, do they actually represent the progression of a natural disease?
Companion animal experiments in the past may have ended with the euthanasia and necropsy of the
animal, but to-day this type of conclusion is not acceptable in the public’s eye.

The validity of trials conducted on dogs and cats kept in a kennel or research facility is questioned, as
these animals do not have the same freedoms and human bonding experiences of the pets kept within a home environment. Most nutritional trials on companion animals are only valid for that particular
group, maintained under the same conditions, fed identical diets. Even the results from the relatively
simple non invasive digestibility, palatability and feeding trials done in kennels or catteries specifically established and approved to conduct these trials have come under scrutiny when environment, previous diet, gender, breed and age differences are considered

 Research methods

The design of the nutritional study determines how significant or relevant the tested diets are in caring for a pet with a specific health problem. The following research designs are listed in order of
importance. (4))4. (http://library.downstate.edu/EBM2/260.htm) These have been modified from the
original to be applicable to veterinary clinical nutrition\. These guidelines provide both the
consumer and the veterinarian with a set of criteria to critically evaluate the research associated with
veterinary medical foods and determine if the product is truly efficacious.

Level 1: Randomized controlled studies

A group of similar animals housed under the same conditions are fed diets, over the same time frame,
one a control diet and the other(s) is(are) intentional modifications of that diet.
 In most trials, the animals are randomly assigned to control, treatment groups’ .Predetermined timed
clinical examinations of the animal, and tests measure the clinical and metabolic response of the animals to these diets. The results are then analyses by what is considered to be the most appropriate statistical program, the results are interpreted and conclusions drawn. Many variations of these trials exist. To eliminate biases the most common are blind studies were the people who are feeding, collecting and analyzing the data are not aware of the status of the diets.

Level 2: Prospective Studies

In a prospective study pets with a specific problem that meets a specific set of clinical signs and
metabolic changes are selected. These patients are divided into several groups and specific diet(s) are
fed. Changes in their clinical and metabolic condition are monitored over a set period using
predetermined clinical and analytical criteria established in the research protocol. A control diet or
control group may be included. These are pets and are not kept for the trial period in a research facility but at home, so patient/owner compliance becomes an issue

 Level3: Retrospective Studies

Retrospective studies are common companion animal nutrition; these studies utilize the owner’s recall of events and the medical records of a group that has a similar condition. The history and records are analyzed to determine if common threads are present within this information that can lead to problem identification, associated risk factors. Conclusions are then drawn and eventually solutions found.
Formally, these are epidemiology studies and are by far the cheapest to do.

Level4: Third person research

A knowledgeable individual or a group of specialists review, interpret and summarize pertinent
published literature on a particular topic. Conclusions drawn and recommendations made. The findings are published in a Journal as a review article or books. The 2006 NRC nutrient requirements for dogs and cats and the AAFCO nutrient profiles for Dogs and Cats used by the industry to define and develop “complete and balanced” diets are examples of this type of research. “Cooperate biases” can strongly influence literature selection.

Most of the industries preliminary formulations and diets were done using this method. A successful
diet or marketing strategy provides the financial resources for further product development.

Level 5: Case-control studies

For these studies, the histories of patients with a certain condition are compared to those without that
problem. In veterinary medicine, the researcher is dependent on the owners to provide a reliable history for both the study and the control group. Once enough cases are collected and similar results found, the researcher may write a review article following a scientifically proven path, design a prospective, or a retrospective study to support the case findings. Randomized control studies are the next level up were the hypothesis are made from the previous studies and tested in a more controlled environment

Level 6: A series a case reports or a single case report

These are generally the first step in developing a research hypothesis they represent a case or a series
of cases that are unique and respond to a particular diet. If these are not published as case reports, they are regarded as antidotal evidence and of limited value. These are considered the lowest level validation

Level7: Educated ideas, editorials, opinions

Although not regarded as scientifically sound, initially most veterinary diets were developed and
marketed based on these criteria.

For Example In The Beginning:
1) Hill’s Pet Food Company (http://www.hillspet.com)
“A company inspired by a guard do"

In 1939, Dr Mark L. Morris Sr. believed certain diseases in pets could be managed through
carefully formulated nutrition

2) Iams Company (http://www.iamsco.com
In 1946, Paul Iams, an animal nutritionist, started the company in a small feed mill near
Dayton Ohio. His aim was to formulate a diet better than that available in grocery stores

3)Nestles Purina (http://www.purina.com)

Founded in1893 by William H Danforth a pioneer in the commercial feed industry, His idea
was that animals must eat year round. In1926, Purina developed diets for the hunting and
working dogs of their rural clients. In 1950, he started to apply the knowledge of farm
animal nutrition to the development of a highly nutritious and palatable dog food sold
through grocery stores. In1957, Purina Dog Chow entered national distribution.

4) Mars Inc. (Master foods) (http://www.mars.com)
“Everyday a new idea”

One of the largest “small family businesses” in the world. Founded by Frank Mars in 1911, who
with his wife Ethel started to sell butter cream candies from their home in Tacoma, Washington.

In 1930, Forrest Mars pioneered the development the European pet food industry combining
modern manufacturing techniques with nutritional science.

Unfortunately the way that these multinational have set up their pet food branches, the consumer still
thinks of them as a small family run business and not as part of a large multinational conglomerate.


Validation of Research

Within the scientific community, none of these research trials is considered valid unless published in a forum that requires peer review and approval prior to publication. This process relies on the integrity and policies of the Journal’s editorial board and the reviewers. To accomplish this can take considerable amount of time.

“Cooperate biases” can weaken the validity of nutritional research by promoting studies with a positive effect and ignoring those showing a negative trend or by amalgamating several weak positive studies produce one strong positive result. Another concern is if the company‘s Research and Development division are interested in supporting a particular conclusion studies showing negative results may-be excluded. Positive trends rather than statistically sound results are cited as positive research in support of a diets formulation and efficacy. In order to properly evaluate companies research the reader must be satisfied that “cooperate bias” did not exist in the formulation or in the marketing of a veterinary medical food. This may-be almost impossible to determine from the information provided by the company.

Confidentiality dictates what research supported by Pet Food Companies is published in peer-reviewed journals. They also bypass this step by holding or sponsoring conferences/symposiums and publishing fact sheets where non-peer reviewed research and the preliminary results of research studies are presented.

Academia

 Let us examine how long this process can take within an academic institution, when non-industry
independent funding is available through grant competitions:

Step 1: Establishing the hypotheses to prove, designing the experiment, getting Animal Care approval, writing and applying for grants to fund the trial can take from 2 to 12 months.

Step 2: The wait for funding approval depends on the deadlines for grant submissions, but the wait can be between 2 to 4 months.

Step 3: Upon approval the trial must be set up this includes establishing the facilities, purchasing the
animals, notifying the testing centers when their services are required, hiring graduate students and
technical help to run the trials. If this can be expedited the time frame could be as short as a month or
as long as four months.

Up to this point, between 5 to 20 months have passed, without a bite of food being eaten.

Step4: The length of actual trial including a period of adjustment depends on the hypotheses but will
likely be between six to 36 months. The ideal period would be over the lifetime of the pet of 10 to 14
years. During this time as the data is being collected, the results received can be collated and
preliminary analyses done.

Step5: All the data is collected, statistically analyzed, interpreted and put into a format suitable for
publication. Depending on how well organized the principle researcher is and the availability of graduate students to do a literature search and writing, this process can take between 3 to12 months to
complete.

Step 6: The actual peer review, manuscript corrections and publication of the research can take
between 6 months to 18 months to complete or the research can be permanently stalled at this step.
One of the major obstacles is when the reviewers submit diametrically opposed corrections and/or
recommendations.

Under ideal conditions, this whole process from step one to six can take from 13 to 66 months to
complete.

The Cooperate Sector

Within the cooperate sector how long this process takes is not as transparent. What we do know or are told is that Research and Development is a big budget item. What we are not told is what percentage of that is allocated to pure scientific research into the development and efficacy of their veterinary products and what is allocated to the research and development into factors that improve product profitability, such as new marketing strategies, new packaging, improved manufacturing methods, flavour development etc. One company has a whole department, headed by an engineer with a PhD, devoted to research into new packaging that is acceptable and convenient for the consumer, maintains the shelf life of the food under the most adverse of conditions and is practical to handle and economical to ship.

All of the companies maintain state of the art companion animal facilities were non-invasive studies
related to diet palatability and digestibilities are done. In some, dogs are trained to assist people with
special needs. Each facility is a showcase and a window through which the public is allowed look and
judge the company. The motives and power that drives the multinational co-operations, and the
marketing strategies used to sell their products are not evident.

The Human Factor

Unfortunately, every step involved in the interpretation, accessing the relevance and evaluating the
quality of research has a major obstacle “the human mind” with all its frailties.

2 comments: