"In this study, we found that Maori and Pacific children received more antibiotic courses than New Zealand European children, as did children living in areas of high socio-economic deprivation compared with those in the least deprived areas," Hobbs said.
"More than one-third of the courses were dispensed during the three winter months and the majority of antibiotic courses were for one antibiotic drug: amoxicillin."
This suggested that many of these courses were likely to have been for seasonal respiratory tract infections which are mostly caused by viruses.
"Antibiotic treatment is not effective in these circumstances and is not supported by New Zealand general practice guidelines," he said.
Antibiotic prescribing decisions for children in general practice often relate to doctors' perceptions of parents' expectations when seeking healthcare, as well as parents' beliefs about antibiotics.
A recent international survey by the World Health Organisation revealed a poor level of public understanding about the situations where it was appropriate to use antibiotics.
Recent research has also shown that the first three years of life is a critical period for establishing a healthy microbiome, the healthy communities of bacteria that live in our gut, on our skin and elsewhere.
It has been suggested that consumption of antibiotics at a young age may interfere with the developing microbiome and lead to permanent changes in immunity or metabolism.
"There is increasing evidence linking antibiotic use in young children to later weight gain and the development of chronic diseases, with asthma being one example," Hobbs said.
"Our finding that more than nine out of 10 children had been exposed to antibiotics by 3 years of age is concerning for this reason."
"The large, ethnically and socio-economically diverse Growing Up in New Zealand cohort gives us an ideal opportunity to investigate this possibility as the children develop.
"We also intend to review general practice medical records to get a better understanding of the reasons for antibiotic prescribing.
"Together, we expect these investigations will provide evidence to help manage the challenges of antibiotic use in primary care including addressing parents' beliefs and expectations and supporting practitioners in balancing benefits versus harms in antibiotic prescribing."
Royal New Zealand College of General Practitioners medical director Dr Richard Medlicott said the findings weren't surprised at the findings, and that the college wanted to see the number of prescriptions reduced over time.
"There is broad recognition within general practice that antibiotics are often prescribed in situation where it's not necessary," Medlicott said.
"The college and other bodies like the Best Practice Advisory Centre have put out various statements and suggestions of ways of trying to reduce usage, but at the same time, being mindful that at some times, they are very useful drugs to use when there is infection that's bacterial and won't get better by itself.
"The difficult is trying to pick which ones need treatment and which ones don't - and at the moment, this evidence suggests that we are over-prescribing."
The study was published in the Journal of Antimicrobial Chemotherapy.
The results
• 97 per cent of the cohort children had been dispensed one or more courses of antibiotics by the time they were five years of age.
• 62 per cent had been dispensed antibiotics in the first year of life with 82 per cent, 92 per cent and 95 per cent in the first two, three and four years, respectively.
• Each child was dispensed a median of eight antibiotic courses by age 5 and the dispensing rate was lower for children living in rural areas compared with urban areas.
• A higher proportion of Maori and Pacific children than European children were dispensed an antibiotic course during each year of life.
• The proportion of children who had been dispensed a course of antibiotics in each month of life increased steadily during the first year, peaking in the twelfth month at one in five children, then declining with age.