This hit the shelves recently (LINK) (for access to the actual article, look HERE, but this option costs money). Recent research followed up on a previous study in mice that saw that mice who received yearly flu vaccines showed less markers for immunity than those that were unvaccinated -- the exact opposite reaction that should hypothetically happen with vaccination.
The human study showed the same results with children.
I have not been pro flu vaccine for... well... most of my existence. The complexity of the flu virus, and how fast it changes, makes it a tricky sucker to predict. Each year there are hundreds of different variations of the virus that circle the globe. Each year scientists try to predict what should go into the national flu regimen based upon what we see circulating around in other countries. Even then, the flu vaccine only contains a few strains (thankfully), when there's hundreds of different strains out there. This all equals a slim chance of getting the proper immunity for the exact strain of flu that hits your household even if the vaccine works as planned (this research points out that it may not).
In almost all studies, immunity acquired when one comes in contact with the virus in usual circumstances (e.g. someone coughing on you, sharing a spoon, etc.) is better than immunity acquired from a vaccination. That makes sense, the immune system didn't develop with vaccinations; it developed in the battlefield of real life -- in dirt, sweat, mingling blood, unwashed hands, freely shared coughs and sneezes in close knit tribal societies.
My opinion on vaccination in general has changed over my career -- there's data showing benefits for vaccinating and not vaccinating. Don't throw the baby out with the bath water, is what I'm trying to say; as well as, think before you inject.