Too much research concentration may have contributed to the erosion of the US' position as a "colossus of science".
A new report by Thomson Reuters shows that although the average citation impact of US-authored papers remains high, its share of papers in top journals has fallen markedly.
In 1981, US scientists produced nearly 40 per cent of papers in the journals indexed by Thomson Reuters, but that had declined to 29 per cent by 2009.
Much of this was due to the rise of the Asia-Pacific region, whose share rose from 13 per cent to 31 per cent: China's share alone rose from 0.4 per cent to 10.9 per cent.
But in the same period, European nations also increased their share from 33 per cent to 36 per cent.
Europe shows a more rounded research base, while the US output contains an "arguably disproportionate" focus on health and social science research, the report says.
The US' highest relative output and citation impact is found in microbiology, biochemistry and clinical medicine, while its share of engineering papers worldwide has fallen 17 percentage points, according to Thomson Reuters' latest Global Research Report.
The average influence of US research papers remains 40 per cent above the world average, but that figure is now rivalled by the UK. And some Asian nations, such as South Korea, China and India, have increased their citation impact "dramatically" in the past decade.
The report also observes that while US research output and excellence has become more concentrated in top institutions, UK research remains relatively well spread.
It suggests that excessive concentration may be damaging US research performance overall.
Responding to "the challenge of agile knowledge economies elsewhere in the world" requires "an equally agile and innovative response, supported by a more pervasive network of US institutions to draw on the talent spread across the 50 states", it claims.
Alan Leshner, chief executive of the American Association for the Advancement of Science, said it was inevitable that increased global investment in the discipline would see the US lose its position as the "sole dominant force".
But he argued that mooted cuts of up to 10 per cent in US federal spending meant the question "isn't so much whether the US is pre-eminent, but whether the eminence of American science will be diminished".
He also agreed that excessive research concentration could be counterproductive: "We benefit from the expanded range of ideas coming from people who don't necessarily think along mainstream, traditional lines."
But Philip Altbach, director of the Center for International Higher Education at Boston College, said similar levels of research concentration were common in other countries. The relatively modest growth of US research productivity was merely "an artefact of a mature academic system with fairly flat research funding", he added.
Stephen Joel Trachtenberg, president emeritus at George Washington University, said there was a "quality of mindlessness" to international comparisons because the whole world benefited from research wherever it was carried out.
"A little bit of national competition may add zest, but academic research is not a foot race," he said.