We recruited 39 volunteers aged 18~23 years old from Shenzhen Polytechnic in 2018. The weight and height of the volunteers were measured by calibrated instruments to calculate body mass index (BMI). All the subjects were given unified physical examinations in professional hospital before and after dietary interventions. Overnight (≥8 hour) fasting blood specimens were collected from each participant for measurement of thyroid hormones, inflammatory, liver and kidney function indicators. Volunteers were included in our study if they fulfilled the following criteria: 1) normal thyroid function (defined as thyroid-stimulating hormone (TSH) of 0.35-5.5 mIU/l, free triiodothyronine (FT3) of 3.5-6.5 pmol/l and free thyroxine (FT4) of 11.5-22.7 pmol/l), no thyroid disorders and treatment history; 2) no exposure to iodine-containing dietary supplement or iodine-containing medication; 3) normal urinary iodine concentration; 4) no functional constipation or diarrhea.

This trial was registered in the Chinese Clinical Trial Registry (ChiCTR1800014877). All procedures involving human participants were approved by the Ethical Committee of the Chinese Center for Disease Control and Prevention. All subjects provided their written informed consent prior to participation.

The sample sizes were calculated by using Gpower 3.1 software. As the study was a before–after self-controlled study, it was estimated that 34 participants would be required for a paired t-test to detect a medium effect size (F=0.5) with a statistical power (1–β error probability) of 80% and an α of 0.05. Thus, our sample size of 37 was adequate.

Considering the potential effect of female physiological period, the whole study was executed limited to a total duration of 18 days, which was divided into acclimation period and three low-iodine intake stages. The project took place in late spring and early summer (March and April) in the southern Chinese metropolis of Shenzhen, with ambient temperatures around 20 degrees. The subjects were required to avoid intensive physical activities during the whole study for controlling iodine excretion from sweat. Diet plans were carefully designed based on balanced meals avoiding from iodine-rich foods and condiments (i.e., kelp, seaweed, iodized salt, etc.). Serving foods and drinking water were restrictively monitored and provide to all the participants. Acclimation period was 6 days before the beginning of the studies in order to reducing the effect of stored iodine in the thyroid (Fig. 1). Subsequently, there were three low-iodine diet (LID) stages, each lasting 4 days (12 days total). In stage 2 and 3, a hard-boiled egg (approx. 50g each) or 125 mL of milk was used for intervention, respectively. The dietary samples, 24-h urine specimens and faeces of volunteers were collected daily by our investigators.

Study design flow chart for 37 Chinese young adults

According to the basic data of the typical dietary pattern obtained from the school canteen and market supply, a 12-day recipe met the principles of nutrient balance was designed by nutritionists. Daily recipe contained staple food, vegetables, fruits, beans and meat, but seaweed and kelp were prohibited. Before food processing, the iodine contents of all cooked foods, including raw materials, compound condiments, snacks and drinking water, were carefully determined. Diets contained different levels of iodine in different periods. All diet cooked with non-iodized salt by the canteen staff. Six drinking water samples collected in school have been detected, and the mean water iodine was 11.8 μg/L. To avoid the bias of iodine content in different drinks, commercially available pure water in the supermarket was used to cook food and drink for subjects. The 12-day recipe used in this study was shown in Supplemental Table 1.

Participants were provided with a plastic beaker (1L) and two plastic bags with sealed lid (5L), and they were guided and demonstrated how to collect the 24-h urine by staff. The 24-h urine was collected for 12 consecutive days from 7 am on the first day to 7 am on the 13th day of the trial. Upon completion of the 24-h collection, the urine volume of each participant was measured and four aliquots were frozen at -20°C for later analysis.

Carmine capsule was administered to all subjects before breakfast to mark feces on the first day, fifth day, ninth day and the thirteenth day respectively. Fecal specimens of each subject were collected according to the labeled red pigmentation at each stage. Once the feces of each stage were completely collected, they were weighed, homogenized, proportionally taken a portion for about 200g and pressed using moulding followed by lyophilization and freezing at -20°C.

During the study, all the diet offered to the subjects was cooked by the university canteen according to the recipe. Trained staff weighed the food supplied for the subjects before and after each meal to access actual intake. A duplicate dietary sample for each meal was collected and homogenized in one hour, and then stored in the refrigerator at -20°C before detection.

Quantitative rapid test kits (Conson Biochemicals, China) were used to detect 24-hour urinary iodine concentration every day to prevent iodine intake other than the diet we provided. Iodine content in food [19], urine [20] and faeces [19] were determined using arsenic-cerium catalytic spectrophotometry at National Iodine Deficiency Disease Reference Laboratory in Beijing. The serum thyrotropin (TSH), free thyroxin (FT4) and free triiodothyronine (FT3) were measured in all participants by an automated chemiluminescence immunoassay analyzer (Bayer ADVIA Cetaur System, Bayer Healthcare, Germany).

Multiple quality control (QC) measures were introduced to guarantee the reliability of the study findings. Prior to the beginning of the survey, all the investigators were trained uniformly for recording weighed food by the principal investigator from National Institute for Nutrition and Health of Chinese CDC. Double entry data have been used to ensure accuracy and completeness. One of the difficulties and limitations in the study was the 24-hour urine samples collection. The total urine volume and 24-hour urinary creatinine excretion were applied to validate the completeness of the 24-hour collection sample. Eating out or consuming food not provided by the project team, was prohibited. The iodine concentrations of 24-hour urine samples were determined promptly every day, and 24-hour urine iodine excretion was calculated to monitor volunteers' dietary iodine intake. Carmine red, a biologically inert substance that was rapidly excreted in faeces, was administered to all participants at the beginnings and ends of each stage to distinguish different stages of stool samples.

Data were analyzed using Excel and IBM SPSS 21.0 Statistics software. The normal distribution of data was checked using the Shapiro-Wilk test. Normally distributed data were expressed as mean ± SD, while non-normally distributed data were expressed as the median (25th-75th percentiles). Differences among the 3 periods were analyzed by one-way repeated measures ANOVA and by the use of Bonferroni correction for multiple comparisons. A P value < 0.05 was considered as statistically significant.

‘Iodine overflow’ hypothesis [18] was first proposed by our group in 2020, which was appropriate for exploring the recommended nutrient intake of iodine under saturation state. Shenzhen, as a coastal city, has implemented USI policy over twenty years and the iodine nutrition status of its residents is appropriate. Therefore, dietary iodine was adequate to meet needs for thyroid hormone synthesis and required storage, while the remaining iodine maybe excreted by ‘overflow’. As we know, iodine excretion increases with iodine intake in certain extent. Dietary contained 3 concentrations of iodine was successively provided for all subjects in 3 periods. Data of iodine excretion increment and intake increment were obtained using the change of iodine intake and excretion between different periods. Ratio was calculated by the iodine excretion increment to iodine intake increment. A scatter diagram was plotted by ratio vs iodine intake, and then the iodine intake was calculated when ratio equaled to 1 with linear regression method.