Should Senior Dogs Be Fed More? The Study That Followed 48 Labradors for Life

Last updated: August 15, 2026

There is a study where the longer-lived dogs were not the ones given more, but the ones not given too much. Labradors raised on the same food at 25% less reached a median age of 13.0 years. The dogs that were not restricted reached 11.2.

48 dogs in pairs; one of each pair was raised for life on 75% of what the other ate
The food was the same. Only the amount changed
Median age reached: 13.0 years versus 11.2 — a gap of 1.8 years (about 16%)
Signs of joint pain arrived on average 1.5 years later
But this is 48 dogs of one breed. It is not about cutting back an already-old dog's meals

She is getting older, so should I switch to something more nourishing? Most people ask themselves this at least once while choosing a senior food.

There is a study that began more than twenty years ago and ran until every dog in it had died. The answer did not point toward adding. It points toward not adding too much.

What gets said

"Now she's a senior, I want to feed her something good." In product copy and in conversations between owners alike, the talk tends to run toward adding. High protein, something for the joints, something for immunity. Almost every word on the shelf is on the adding side.

Cutting back only ever comes up as weight loss for an overweight dog. For a dog at a healthy weight, the idea of keeping the amount modest barely gets discussed.

How the study was built

Forty-eight Labrador retrievers were organised into pairs. From eight weeks of age, one of each pair was fed as much as it wanted, and the other was given 75% of that amount. This continued until the dogs died.

The only thing that changed was the amount. The same food, same ingredients, same formulation. Nothing in the design lets the effect of adding something get mixed up with the effect of feeding less.

This design is rare in food research. Most comparisons look at groups eating different products, so ingredients, formulation, kibble size, and palatability all differ at once. Whether the study narrowed its variable to one is among the first things worth checking as you read.

The restricted dogs were not going hungry. Because the partner ate freely, the design is less "a lean group versus a normal group" and closer to a group kept from getting heavy versus a group fed freely.

And it ran for life. Not a trial of a few months, but a record kept until every dog had died. Lifelong follow-up on this scale is almost unheard of in dogs, which is why the study is still being cited.

The numbers

Comparison of 48 Labradors (reported 2002)
Measure	Fed 25% less	Fed freely
Age half the dogs reached	13.0 years	11.2 years
Signs of joint pain	On average 1.5 years later	—
Onset of chronic disease	Later	—

1.8 years, about 16%

The age half the dogs reached was 13.0 years for the restricted group and 11.2 years for the freely fed group. The gap is 1.8 years, or about 16%.

A median is the value sitting exactly in the middle when everything is lined up in order. Here it means the age half the dogs reached. A mean gets pulled by one unusually short- or long-lived animal; a median is far less sensitive to that.

It was not only lifespan that stretched

Signs of joint pain appeared on average 1.5 years later, and chronic disease started later too. This is not a case of living longer but spending that time unwell.

When people hear that lifespan increased, some think it just means a longer stretch of nursing care. In this study at least, it did not. The healthy period stretched along with the total.

What is not known

First, this is 48 dogs of one breed. What happened in Labradors will not necessarily apply to every dog by the same margin. Body size is known to affect the pace of aging as well (large dogs and the rate of aging).

And this ran from eight weeks of age for the whole of life. It is not grounds for cutting an already-old dog's meals starting today. In the senior years, the problem is sometimes not losing weight but losing too much. Reducing the amount for a dog whose appetite has been fading is a different subject from what this study addresses.

One more thing: this study controlled the amount given. It did not rank particular foods. There is no answer here about which food is better.

Why feeding less extended life is outside the study's scope. Whether it runs through differences in weight and body fat, or through some other route, remains a separate question.

What this study is good for

If you live with a young or middle-aged dog

It is a reason to check the feeding guide against your dog's body shape now and then. Not letting a dog get heavy while young may be what shows up as a difference in later years — that is about the size of the claim.

Feeding guides are usually written as a range, and the amount changes considerably depending on whether you sit near the top or the bottom of it. For a dog that does not get much exercise, starting toward the lower end and adjusting by body shape runs in the same direction as this study.

If your dog is already a senior

There is no need to change the amount on the strength of this study. Whether reducing or increasing, weight trends and your own vet's read come first.

What the study offers is that "add something good" is not the only way to think about food. It is something to pause on, standing in front of the shelf, if every word in front of you is on the adding side.

A fading appetite is a different subject

This study does not apply to a dog who has been eating less. Posture and the state of the mouth can both be involved in not eating. The order worth trying things in is on the page about a pet who will not eat.

How this number gets misread

Misreading 1: "Cut meals by 25% and gain 1.8 years"

This is the result of a regimen run from eight weeks of age for life. What happens if you cut back partway through was not tested.

In the senior years especially, losing too much weight is sometimes the greater problem. Reducing the amount for a dog whose appetite is fading can run in the opposite direction from what this study found.

Misreading 2: "Eating less is healthier"

The comparison was "75% of the partner" against "as much as it wants." Rather than a small portion versus a normal one, it is closer to a group kept from getting heavy versus a group fed freely.

Mechanically subtracting 25% from the feeding guide on the package is not a practice this study endorses.

Misreading 3: "This food makes dogs live longer"

Both groups ate the same food. This study says nothing about which food is better. If it turns up in a product description, that part needs separating out as you read.

Misreading 4: "Large dogs live longer if you keep their weight down"

Two very different studies tend to get mixed together in talk about weight and lifespan. The finding of about one month per 2kg came from comparing breeds against each other in a separate study, which is not about varying the amount within the same dogs.

Both look like statements about weight, but they answer different questions. Mixing them distorts both conclusions.

Common questions

If I cut back my senior dog's food now, will she live longer?

That is not what this study says. The amount was controlled from eight weeks of age through the whole of life; it did not test cutting back once a dog is already old. In the senior years, losing too much weight is sometimes the greater problem. If you are going to change the amount, weight trends and your own vet's read come first.

Does it tell me which food is better?

No. This study used the same food and varied only the amount. It was not comparing brands or formulations, so it cannot be used to justify a product choice.

Does "25% less" mean 25% off what the bag says?

No. It means 75% of what the paired dog ate. That partner ate as much as it wanted, so in practice the design is closer to a comparison between a group kept from getting heavy and a group fed freely. It is not about subtracting 25% from the feeding guide on the package.

Is the 1.8 years an average lifespan?

It is the age half the dogs reached — a median. The figures were 13.0 years and 11.2 years, a difference of 1.8 years, or about 16%. A median is calculated differently from a mean.

Does the same hold for other breeds?

This study alone cannot say. The subjects were 48 Labrador retrievers. A result from a limited number of dogs in a single breed does not necessarily apply to all dogs by the same margin.

Sources

Kealy RD, Lawler DF, Ballam JM, et al. Effects of diet restriction on life span and age-related changes in dogs. J Am Vet Med Assoc. 2002;220(9):1315-20. PMID: 11991408 (figures cross-checked against the abstract and text, and against material published by the company and university that ran the study; accessed August 14, 2026)