Columbia University Medical Center
In Vivo - The Newsletter of Columbia University Medical Center
Back Issues
Contact Us
In Vivo

Prenatal Air Pollution Exposure Seems to Slow Cognitive Development
Children whose mothers were exposed to high levels of air pollution during pregnancy scored significantly lower on mental development tests than other children, say researchers in the Columbia Center for Children’s Environmental Health.
    The research adds to an increasing body of evidence that shows environmental pollution has a significant impact on children’s health.
    The study examined 183 3-year-olds born to nonsmoking mothers who carried personal air monitors during their third trimester of pregnancy. The monitors collected polycyclic aromatic hydrocarbons (PAHs), which are released into the air during combustion from cars and trucks, power plants, and cigarettes, among other sources.
    Children whose mothers were exposed to the top quartile of PAH levels during the study Prenatal Air Pollutionscored on average 5.7 points lower on cognitive tests than the other children. Their odds of being classified as moderately delayed were also almost three times higher than among less exposed children, which indicates they are more likely to perform poorly in the early school years.
    The magnitude of the effect is comparable to that reported by others for prenatal low-level lead exposure.
    The findings are the first to show an effect of prenatal PAH exposure on child neurodevelopment, says the study’s lead author, Frederica Perera, Dr.P.H., director of the center and professor of environmental health sciences in the Mailman School of Public Health, and need to be confirmed by other studies. However, the findings are consistent with experimental studies on PAHs and with prior human studies implicating air pollution more generally; they support initiatives to reduce fossil fuel emissions in the urban environment.
Environ Health Perspect, in press, doi:10.1289/ehp.9084 available via http:// The research was supported in part by the NIEHS and EPA.

Finding Patients for Research Difficult in Privacy Era
When neurosurgeons at Columbia tried to set up the first tissue bank for a newly described pediatric brain tumor, they found themselves tangled in a complex web of privacy regulations created by the Health Insurance Portability and Accountability Act (HIPAA).
    Under HIPAA’s Privacy Rule, researchers who want protected health information from patients must obtain authorization from each patient. But in interpreting the rule, Columbia, like most institutions, has forbidden researchers from contacting these patients unless their treating physician first gives written permission.
    “Because we’re not the treating physicians for some of the children with this tumor, the privacy rule makes it extremely difficult to enroll these patients,” says Ricardo Komotar, M.D., a postdoctoral residency fellow in neurosurgery. “It has really hampered our ability to start our research into the underlying genetic mutations that cause this tumor.”
    To boost recruitment, Dr. Komotar, Richard C.E. Anderson, M.D., assistant professor of neurosurgery, and others have set up a Web site that takes advantage of the Internet’s popularity as a source for health information. Parents get accurate information at the site and are also asked to donate samples to the tissue bank. With the patient’s consent, the neurosurgeons are then able to contact the treating physician for clinical information and tissue samples. Once obtained, any information that could identify the patient is stripped from the data so other researchers who want to use the tissue bank do not have to obtain additional consents.
    Since it went live in August 2004, the site has helped enroll 10 patients, including three international patients, into the tissue bank. Given the rarity of this tumor, the researchers say this number is outstanding.
    “We believe that this approach may be helpful for physicians in other disciplines as well, and we hope that Internet recruitment will make it easier for health care professionals to conduct meaningful research,” Dr. Anderson says.
Neurosurgery 58(5): 985-989. The work was supported by the NIH, Congress of Neurological Surgeons, and a Doris Duke Clinical Research Fellowship.

Lack of Sleep Linked to Hypertension
A lack of sleep has been linked to poor health, including obesity, diabetes and, in women, heart attacks. Now, the latest study on the dangers of drowsiness shows that middle-aged people who sleep five hours a night or less also have a two-fold higher incidence of hypertension than their more well-rested peers.
    The study examined data from 3,620 people aged 32 to 59 who had normal blood pressure at the time they participated in the first National Health and Nutrition Examination Survey in the early 1980s.
    Ten years later, 24 percent (57 out of 241) of people who reported sleeping five hours a night or less had developed high blood pressure compared to only 12 percent (290 out of 2451) who reported sleeping 7 to 8 hours.
“Sleep allows blood pressure to drop, so people who sleep little raise their average 24-hour blood pressure,” says the study’s lead author, James Gangwisch, Ph.D., post-doctoral fellow in the psychiatric epidemiology training program at the Mailman School of Public Health. “The higher average may set up the cardiovascular system to operate at an elevated pressure.”
    If this is true, Dr. Gangwisch adds, interventions to improve sleep could serve as treatments and as a primary preventative measure for hypertension.
Hypertension 47: 833-839. The research was supported by an NIMH award to the Psychiatric Epidemiology Research Training Program.

Researchers Probe World of Non Coding RNA
When the Human Genome Project was completed and scientists started counting the number of genes, most expected the genome to contain close to 100,000 genes. The latest tally, however, puts the number at around 25,000, about the same as the C. elegans flatworm. Certainly humans are more complex than flatworms, but how do we get that way with essentially the same genome?
    Many now think complexity may stem from an organism’s non coding RNAs, which are spun off the genome but not translated into proteins. The extent of this hidden RNA world – which may comprise 98 percent of human genomic output – has only begun to be appreciated in the last few years. As interest grows in this “dark matter,” it is increasingly important to have a method that distinguishes between non coding RNAs and “old-fashioned” messenger RNAs, which are translated into proteins.
    Researchers in Columbia’s Center for Computational Biology and Bioinformatics have now created the most accurate method yet to identify non coding RNA. Employing a computer program that was taught to distinguish between protein-coding and non coding sequences, the technique identified about 14,000 different non coding RNAs in the mouse genome, say Burkhard Rost, Ph.D., associate professor, and Jinfeng Liu, Ph.D., associate research scientist, Department of Biochemistry and Molecular Biophysics. Based on the technique’s error rate, they also estimated approximately 14,000 more non coding RNAs are in the mouse transcriptome.
    “Certainly 28,000 is much more than we anticipated,” Dr. Rost says. “It’s similar in magnitude to the number of proteins, but unlike proteins, we only know what a tiny fraction of these non coding RNAs are doing.” Dr. Rost adds that the 28,000 number also underestimates the non coding RNA world, because the program focuses on long RNAs (more than 80 nucleotides), omitting short non coding RNAs (such as microRNAs) from the analysis.
    Drs. Rost and Liu say the program will be available in the future on their Web site list so other researchers can submit their own sequences and identify the non
coding RNAs.
PLoS Genetics 2(4): 529-536.
The research was supported by the National Library of Medicine and the Japanese Ministry of Education, Culture, Sports, Science and Technology.