*3.1. PA Concentrations in (Herbal) Teas*

In this study, a total of 68 individual PAs were analyzed, of which 23 PAs were not detected in all the investigated (herbal) teas (supporting data S2). Seventeen PAs were found in lemon balm (melissa) tea, ranked first regardless of the REP correction, followed by tephroseris and lemon balm & liquorice (Table 2). None of the targeted PAs were present in citroen melisse and fresh peppermint tea. Most of the teas (92.5%) were detected with PAs, with the measured total levels varying from 13.4 μg/kg d.m. to 286,682.2 μg/kg d.m. (Table 2). In terms of the total PA content, tephroseris, borage and lungwort were the top three teas, which all originated from PA-producing plants. When taking the REP factors into account, the concentrations of total PAs were ranged from 1.3 μg/kg d.m. to 286,648.3 μg/kg d.m., which were generally lower compared to those measured levels in the (herbal) teas, except for green tea, gynura segetum and rooibos. It is of note that the total PA level for borage tea decreased by about 116.5-fold by the REP correction, amounting to 1440.6 μg/kg d.m. Whereas the PA concentrations in tephroseris were hardly altered by the REP correction, lungwort presented a considerable drop in its total PA level. This drop made the REP-corrected PA level for lungwort even lower than that for some teas from non-PA-generating plants, such as lemon balm, chamomile, rooibos and lemon verbena. Among the teas derived from non-PA-producing plants, lemon balm (melissa) and chamomile ranked first and second with their total PA levels, irrespective of REP correction.

The regulations of Germany and the Netherlands have indicated that the maximum limit for daily intake of 1,2-unsaturated PAs (including *N*-oxides) during a lifetime by a 70 kg person was 0.1 μg/day [24,25]. In this study, we calculated the daily intake of total PAs by consuming one cup of tea with and without REP correction. The results showed that there were nine types of teas resulting in the daily intake of PAs above 0.1 μg/day, regardless of REP correction (Table 2). These teas included tephroseris, borage, lemon balm (melissa), chamomile, eupatorium, rooibos, mix herb (1), lemon verbena and green tea. In addition, the daily intake of PAs due to the consumption of earl grey, lemon balm & liquorice, lungwort and sage & lemon myrtle could exceed the maximum limit when the REP factors were not applied.

According to the top three PAs and their concentrations, senkirkine and its congener neosenkirkine were the dominant PAs in the (herbal) teas from PA-containing plants except for borage and lungwort, while echinatine, retrorsine, integerrimine and senecionine as well as their *N*-ox congeners were frequently occurring in the non-PA-producing teas (Table 2). When taking the REP factors into account, senkirkine, neosenkirkine and petasitenine remained the same levels as their REP factors were derived to be 1 (supporting data S1, Table S2). In contrast, supinine, intermedine and their *N-*oxide congeners plus lycopsamine *N-*oxide correspond with a proposed REP value of 0.01, which could explain for the remarkable decreases in the total PA levels for borage and lungwort due to REP correction. Overall, the types of the top three PAs remained unchanged in most of the studied teas in response to REP correction, with one exception of lemon balm & liquorice, which was due to the REP factors of atropine and scopolamine, set at 0.

*Foods* **2022**, *11*, 2946

of

**Table 2.** Total number of detected PAs, total PA levels, daily intake of total PAs when consuming 2 g of tea and the top three PAs and their concentration in 21 types (herbal) teas with and without REP correction. A complete data overview is shown in supporting data S2.


*Foods* **2022**, *11*, 2946


in Italic while those with both values in bold are underlined. a Only two types of PAs were detected out in forest fruit tea and gynura segetum tea.
