Preparatory work for the update of the tolerable upper intake levels for vitamin A

This report describes collection of data and appraisal of scientific evidence in relation to assessment of Upper Levels (UL) for vitamin A. In the SR on teratogenicity, 370 records were identified and reduced to 23 records after screening (3 prospective and 20 case‐control studies). Few studies reported intakes in the range of 3,000 ‐ 15,000 μg per day. There is considerable heterogeneity in outcomes. Available body of evidence (BoE) demonstrates no overall trend of increased risk of teratogenicity with exposure to high preformed vitamin A or total vitamin A (preformed vitamin A and β‐carotene). In the SR on hepatotoxicity, 918 records were identified and reduced to 4 records (randomized controlled trials) after screening. The studies examined doses ranging from 300 to 41,250 μg RE/day of vitamin A. There was no clear overall trend of increased risk of hepatotoxicity with exposure to high vitamin A. Available BoE demonstrates lack of scientific data to establish a dose‐response relationship. In the SR on bone fractures, 795 records were identified and reduced to 11 records (prospective studies) after screening. Meta‐analysis was performed on hip fractures in women. There were limited/suggestive evidence for increased risk of fractures with preformed and total vitamin A intakes >1,500 and > 3,000 μg RE/d, respectively. However, available BoE is insufficient to conclude that high vitamin A increases risk of bone fractures in humans. In the SR on bone mineral density (BMD), 795 records were identified and reduced to 7 records (prospective cohort studies) after screening. There was considerable uncertainty related to source of vitamin A. There were limited/suggestive evidence for increased loss of BMD with higher intake of preformed vitamin A from diet including supplements in elderly women (data only available in one study). However, the available BoE is insufficient to conclude that high vitamin A affect BMD in humans.