This study recruited 23 North Korean refugees from a religious facility in Guro-gu, Seoul. The participants were aged between 18 years and 34 years old and provided informed consent before the assessments. The study protocol was approved by the Institutional Review Board (IRB SJU-HR-E-2019-008) of Sejong University, Seoul, Korea. We measured the height and weight of the subjects, and obtained information on their age, cigarette smoking habits, and consumption of alcoholic beverages. During face-to-face interviews, we asked the North Korean refugees for the date of their escape from North Korea, the date of their arrival in South Korea, and the number of months they spent in transit between the two countries (i.e., in China) to calculate the length of their residence in South Korea. Heavy smokers (≥20 cigarettes per day) and heavy alcohol consumers (≥5 drinks of alcohol per day; one drink is equivalent to 12 g of alcohol) were excluded from the analysis because these substances can modify taste perception [8]. We also excluded those who could not distinguish between distilled water and the sample solutions. Ultimately, 21 young adult North Korean refugees (12 men, 9 women) participated.

In total, 72 young adult South Koreans (40 men, 32 women; age range: 18-26 years) were recruited from universities in the metropolitan area of South Korea; they all provided informed consent before participating. The subjects were asked the same questions as the North Korean refugees, and the exclusion criteria were identical. Finally, 41 South Korean young adults (20 men, 21 women) were included in the analysis. Table 1 summarizes the general characteristics of the participants.

To determine the taste stimuli detection and recognition thresholds, we prepared a series of solutions corresponding to the four basic tastes (sweet, salty, sour, and bitter) using sucrose, sodium chloride (NaCl), citric acid, and quinine hydrochloride (QHCl), respectively. The concentration range for each test compound was chosen based on the reported threshold values. The highest-concentration solutions were 40% sucrose, 20% NaCl, 2% citric acid, and 0.1% QHCl. Solutions with lower concentrations were obtained by serial dilution with distilled water (ratio of 1:1), as shown in Table 2. All test solutions were stored at 4°C until use.

To ensure the accuracy of the threshold tests, the subjects were instructed to refrain from eating a meal, drinking (except water), smoking, or chewing gum for at least 1 h before the start of the test. Before the test, the subjects rinsed out their mouths with distilled water; then 1 mL of each stepwise test solution was sprayed evenly over the subject’s tongue surface using a disposable syringe. The subjects retained each solution in their mouths for a short period and then spat it out. Each subject was asked first to discriminate between the taste of the test solution and the previous solution, and then to identify the taste of the test solution. The minimum concentrations at which the subjects could discriminate the test solutions from distilled water, and identify the taste correctly, were defined as the detection and recognition thresholds, respectively. When subjects switched from one test solution to another, they first rinsed out their mouths with distilled water. However, when the concentration of the same test solution was increased, they did not rinse their mouths, except in the case of the salty taste.

Data are presented as means±standard error of the mean. Statistical significance of difference was analyzed by chisquare test on gender and one-tailed Student’s t-test on other variables. Significant differences were accepted when the p value was <0.05.

Two different gustatory thresholds are often used to assess taste sensitivity in humans: the recognition and detection thresholds. The detection threshold is defined as the minimum concentration of a taste chemical that can be reliably discriminated from water, although the specific taste is not necessarily recognized. The recognition threshold is defined as the minimum concentration at which a subject can accurately identify the type of taste, such as “sweet” for sucrose or “salty” for NaCl [9,10].

Fig. 1 shows the cumulative curves of the recognition thresholds and mean recognition thresholds for the sweet, bitter, sour, and salty stimuli in our young adult North Korean refugees and South Korean participants. In the cumulative curves of the recognition threshold, the onset of the curve corresponds to the lowest threshold and the curve is plotted up to the highest threshold; the steepness of the slope is inversely related to the range width of threshold for each group. As shown in Fig. 1A, the onset, end point, and steepness of the cumulative curves of the recognition threshold for QHCl were quite similar between the North Korean refugees and the South Koreans. No difference was observed in mean recognition threshold between the two groups (1.0024 mg% vs. 0.9359 mg%, p=0.394). Regarding the sensitivity to citric acid, both the cumulative curve of the recognition threshold and the mean recognition threshold (0.0201% vs. 0.0189%, p=0.382) showed no differences between the groups (Fig. 1B). These data might suggest that a sociological factor such as a prolonged food deficit does not strongly influence the recognition thresholds for bitterness and sourness in North Korean refugees. On the other hand, different sucrose sensitivity trends were found between the two groups (Fig. 1C). As expected, the cumulative curve of the recognition threshold for sucrose shifted to the left for the North Korean refugees compared with the South Korean participants. The mean recognition threshold for sucrose was also significantly lower in the North Korean refugees (0.5357% vs. 0.7393%, p=0.044), indicating that sucrose sensitivity is greater in North Korean refugees than in South Korean participants. Similarly, we found a difference in taste sensitivity between the groups for NaCl in terms of both the cumulative curve of the recognition threshold and mean recognition threshold (Fig. 1D). The cumulative curve of the recognition threshold among the North Korean refugees shifted to the left for NaCl concentrations <0.03906% and to the right for NaCl concentrations >0.03906% compared with the South Korean participants. The mean recognition threshold of the North Korean refugees was significantly higher than that of the South Korean participants (0.2174% vs. 0.1212%, p=0.027), indicating that sensitivity to NaCl was lower in the North Korean refugees than in South Korean participants.

Fig. 2 shows the cumulative curves of detection thresholds and mean detection thresholds for the sweet, bitter, sour, and salty stimuli for the two groups. As with the recognition threshold, the onset of the curve corresponds to the lowest threshold, and the curve is plotted up to the highest threshold; the steepness of the slope is inversely related to the threshold range width for each group. Unlike recognition thresholds, increasing the concentration of all taste stimuli tested produced a rightward shift of the cumulative curves of detection thresholds among the North Korean refugees. The mean detection thresholds for sweet (0.2400% vs. 0.4059%, p=0.038), salty (0.0188% vs. 0.0373%, p=0.002), sour (0.0030% vs. 0.0100%, p=0.001), and bitter (0.1171 mg% vs. 0.4595 mg%, p=0.0005) tastes were significantly lower in North Korean refugees than in South Korean participants, indicating that the overall sensitivity of detection to taste stimuli was higher in North Korean refugees than in South Korean participants.