The number of people living with HIV in India could be lower than government estimates, research published in a medical journal said, but the United Nations warned against drawing "hasty conclusions".
Scientists who studied the prevalence of the virus that causes AIDS in a district in Andhra Pradesh -- the state has the highest HIV rate in India -- found it was less than half the government's figure.
Instead of 112,600 HIV cases in the district of Guntur, Lalit Dandona, of the Administrative Staff College of India, in Hyderabad estimates the number is about 45,900.
"The official method in India leads to a gross overestimation of the HIV burden in this district," Dandona said in a report published in the journal BMC Medicine.
"The potential major implications of these findings for the overall HIV estimate for India need to be examined," he said.
The United Nations estimates India has the largest number of people living with HIV/AIDS at 5.7 million and its anti-AIDS agency warned against extrapolating conclusions about the spread of HIV in the country from the Guntur study.
India calculates the HIV caseload through surveillance in selected clinics and hospitals around the country.
The sample size means the surveillance data from one particular district may not necessarily give an accurate picture of number of cases in the district itself. But in aggregate UNAIDS believes the data from the whole country gives a fair estimate.
"Even if we could find there is an overestimation in Guntur, it is not acceptable to conclude this applies to the whole of India," UNAIDS India chief Denis Broun said.
"One should not jump to conclusions too hastily," Broun said.
Nevertheless Broun said the Hyderabad staff college was highly respected by UNAIDS and other agencies.
Andhra Pradesh has the highest incident of HIV-positive people in India with over two per cent of the population or nearly 1.5 million people living with the virus.
The researchers of the report, published in the BMC Medicine journal, tested blood samples from more than 12,000 men and women aged 15-49 from both urban and rural areas in Guntur who were representative for the study.
The HIV prevalence rate they found was 1.72 per cent and rose to 1.79 per cent, or 45,900 cases, after they adjusted the number for high risk groups.
Official figures based on data collected from antenatal clinics, sexual health clinics, high risk groups and referrals of HIV positive and suspected cases to public hospitals put the number of cases at 112,600.
Dandona said public hospitals have disproportionately high rates of HIV because they are used more by patients from lower socioeconomic groups who have a higher prevalence of HIV.
UNAIDS's Broun said population survey results often gave different results from other methods such as surveillance site data, but added the global body would review its current estimates if needed.
In the past, UNAIDS figures for HIV-positive people in East Africa, including Kenya, were found to be overestimated.