LOS ANGELES -- The steady decline in reported deaths caused by alcohol, one of the happiest social and medical trends of the 1980s, has stopped nationwide and been reversed in California, health statisticians report.

The change has caused concern among doctors and alcohol abuse workers, who express puzzlement after what appears to have been a decade-long decline in drinking by most Americans, including Californians.

Among possible causes, researchers said, are the new popularity of wine coolers, an alcoholic drink often marketed like a soft drink to young adults, and rebounding sales of other alcoholic beverages.

Advertising is encouraging young people to move "from non-alcoholic beverages to wine coolers," said Christine Lubinski of the National Council on Alcoholism and Drug Dependence. Beer sales climbed 3 percent in 1989, the first increase in at least three years.

Lubinski and several other experts suggested that the new increase in reported deaths also may reflect a growing willingness to acknowledge alcohol as a cause of death. Alcoholism was "probably the most underreported cause of death until the advent of the AIDS tragedy," said Chauncey Veatch, director of the California Department of Alcohol and Drug Programs.

Physicians are better trained to recognize symptoms of alcohol abuse and more likely to note that on death certificates despite family embarrassment, several experts said. Gary Levine, a pediatrician in Mill Valley, Calif., who specializes in alcohol abuse, said doctors now will say "that Aunt Sadie died because she was drinking excessively" and report that to agencies that compile death statistics.

The reversal in California's alcohol death rate was revealed in a new report by James W. Sutocky of the state Department of Health Services. It showed that, although the alcohol death rate -- not including drunken driving deaths -- declined from 17.3 per 100,000 population in 1979 to 13.3 in 1985, it had climbed to 14.8 by 1988, the last year for which figures are available.

Sandy Smith, spokeswoman for the National Center for Health Statistics in Hyattsville, Md., said national statistics show a similar, if much smaller, increase in the alcohol death rate since 1986 after a marked decrease in the early part of the '80s.

The new death rate figures come at a time of renewed clashes between alcohol abuse experts and the liquor industry. Surgeon General Antonia C. Novello has sharply criticized the company selling Cisco, a fruit-flavored wine marketed like a wine cooler but much more potent.

Doctors reported that 10 of 15 adolescents admitted to Children's National Medical Center in Washington with acute alcohol poisoning from March through December said they had been drinking Cisco.

Controversy about alcohol advertising, including use of young actors and youthful themes such as carefree trips to the Super Bowl, has led the Washington State Medical Association to propose what would be the nation's strictest advertising guidelines.

The proposal, to be considered by the state liquor control board Wednesday, would ban advertisements linking alcohol consumption with "athletic prowess or professional or social achievement of any kind" and ban use of themes or objects, such as toys, attractive to children.

Annual American liquor consumption measured by tax receipts and industry reports peaked at 2.76 gallons of pure alcohol per person in 1980 and 1981 and has sharply declined since then. By 1987, according to the National Institute of Alcohol Abuse and Alcoholism (NIAAA), consumption had dropped to 2.54 gallons, lowest since 1970.

Consumption in California dropped from 3.25 gallons in 1977 to 3.12 gallons in 1986. Consumption in Maryland between those years also declined, from 3.05 to 2.76 gallons, but in the District, which has the nation's highest rate, consumption increased from 5.53 to 5.67 gallons.

Analysts said the District's rate probably was inflated by its lower alcohol taxes, which attracted buyers from Maryland and Virginia, and its tourist industry.

Virginia's consumption rate grew from 2.3 to 2.53 gallons, at 10 percent the largest increase in the country.

Causes of death included in mortality rate statistics released by California were alcoholic psychoses, dependence syndrome, nondependent use, hepatitis, cirrhosis of the liver and other damage, heart diseases, excess blood alcohol level and alcoholic poisonings.

Drunken driving fatalities were not included in the multiyear study, although California authorities reported a 15.5 percent increase in drunk-driving arrests in the first 10 months of 1990.

Mary Dufour, a physician and chief of the NIAAA's epidemiology branch, cautioned against drawing immediate conclusions from the reversal in death rates. For decades, she said, researchers have seen a close connection between death rates from cirrhosis of the liver and alcohol consumption but still cannot explain why the cirrhosis death rate began to decline in 1973, almost a decade before Americans started drinking less.