The fist of its kind study by a team at the Emory University, Atlanta, found that teens who eat elevated amounts of added sugars in drinks and foods are more likely to have poor cholesterol and triglyceride profiles, which may lead to heart disease later in life.
The study, published in the journal Circulation, also showed that overweight or obese teens with the highest levels of sugar intake had increased signs of insulin resistance, often a precursor to diabetes.
According to the American Heart Association, "added sugars" are any caloric sweeteners added to foods or beverages in the manufacturing process or by the consumer.
"Adolescents are eating 20 per cent of their daily calories in sugars that provide few if any other nutrients," said Jean Welsh, study author and post-doctoral fellow in paediatric nutrition at Emory University School of Medicine.
"We know from previous studies the biggest contributors of added sugars to the diet are sugar-sweetened beverages such as sodas, fruit-flavoured drinks, sweetened coffees and teas".
For the study, the first study to assess the association of added sugars and the indicators of heart disease risk in adolescents, Welsh and colleagues looked at 2,157 teenagers aged between 12 and 18 years.
They found that the average daily consumption of added sugars was 119 grams (28.3 teaspoons or 476 calories), accounting for 21.4 per cent of their total energy.
Teens consuming the highest levels of added sugars had lower levels of high-density lipoprotein levels (HDL) -- the good cholesterol -- and higher levels of triglycerides and low density lipoproteins (LDL)-- the bad cholesterol.
Teenagers with highest levels of sugar intake at over 30 per cent of total energy had 49.5 milligrams/deciliter (mg/dL) compared to 54 mg/dL of HDL levels in those with the lowest levels of added sugar consumption -- a 9 per cent difference.
The study included dietary recall from one 24-hour period that researchers merged with sugar content data from the US Department of Agriculture My Pyramid Equivalents Databases.