此台灣第二次總膳食調查（TDS）計畫之目的，在於評估一般消費者
攝入食品中，二項防腐劑苯甲酸及已二稀酸之健康風險。藉此，重新強化
我國TDS架構，以便與國際食品安全管控主流接軌。本計畫之健康風險，
乃以分析標的物之每日暴露量對其每日耐受量（ADI）之比例來評估。按
定義，此比例稱為危害指數（HI），必須小於100%，為食品安全之要求。
評估所需之ADI數值，取自聯合國食品添加物聯合專家委員會（JECFA）
公布資料。由於只有苯甲酸及已二烯酸頒有ADI建議數值，故本計畫僅
聚焦於此兩類防腐劑。每日暴露量之計算需用到三套數據：1）食品中標
的物之濃度（C） 、2）各項食品之個別攝食量（CR） 、及3）消費者之體重
（BW）。C數據由化學分析檢測食品中標的物而得；CR 及BW數值則來自
國民營養健康狀況變遷調查計畫數據檔，經加工分析而得。HI 之數值由下
列綜合公式計算而來，結果以%ADI呈現。
2

of these food samples, and 3) the body weight (BW) of consumers. C values were
obtained by chemical analysis of the analytes in food samples, while CR and BW
values were derived from the database of the Nutrition and Health Survey in Taiwan
(NAHSIT). The HI values were calculated using the equation below and expressed as
% ADI.
100% ADI BW
CR C
100% ADI
ADD
%ADI HI
n
1 i P j
ij i
n
1 i P
ij
n
1 i
ij j
×
×
×
= × = = ∑ ∑ ∑ = = =

A scheme was developed to select food items to represent the total diet
consumed in Taiwan. Based on the NAHSIT database, the total diet was classified
挑選食品項目用以代表台灣日用總膳食之設計如下，根據NAHSIT數
據檔，台灣總膳食依營養成分可分為12大類、47小類；再依攝食量大小排
序，選出128細項作為TDS基礎清單，代表了台灣日用總膳食量之83%。
然後考慮標的物之實際分布情況，進而把基礎清單調適為97項食品之防腐
劑專屬清單，作為採樣及化學分析用食品樣本製備之依據。樣本食品之採
買分春夏及秋冬兩季，連續兩年執行於全國50個策劃點。食品樣本依入口
狀況製備，經勻質和混合後送檢以產出濃度（C）數據。消費者九個年齡層
之體重（BW）及各項食品之個別攝食量（CR），則自NAHSIT數據檔分析
得來，用為計算風險所需之暴露參數。風險計算結果顯示，一般國人之苯甲酸暴露量低於50%ADI 、已二烯酸暴露量約在10%ADI左右。此結果表明
當前台灣總膳食中所含苯甲酸及已二烯酸，沒有導致健康危害之虞。本計
畫之完成，確已重新強化了我國之TDS架構和量能，有利於針對更多食品
中重要標的化學物，永續執行風險評估。

This second total diet study (TDS) in Taiwan was initiated to assess the health risk posed on general consumers from the dietary intake
of benzoic acid and sorbic acid as an illustration of a risk assessment of food preservatives. Health risk was assessed by estimating the ratio
of the exposure level to the acceptable daily intake (ADI) level of an analyte. This ratio, known as the hazard index (HI), must be less than
100% to ensure no health risk of concern. The ADI values were obtained from the Joint FAO/WHO Expert Committee on Food Additives
(JECFA). The present TDS was conducted based on the two analytes, benzoic acid and sorbic acid. Exposure levels of the analytes were
determined by using three sets of local data: 1) the analyte concentrations (C) in pertinent food samples, 2) the individual consumption
rates (CR) of these food samples, and 3) the body weight (BW) of consumers. C values were obtained by chemical analysis of the analytes
in food samples, while CR and BW values were derived from the database of the Nutrition and Health Survey in Taiwan (NAHSIT). The
HI values were calculated using the equation below and expressed as % ADI.

2

obtained by chemical analysis of the analytes in food samples, while CR and BW
values were derived from the database of the Nutrition and Health Survey in Taiwan
(NAHSIT). The HI values were calculated using the equation below and expressed as
% ADI.
100% ADI BW
CR C
100% ADI
ADD
%ADI HI
n
1 i P j
ij i
n
1 i P
ij
n
1 i
ij j
×
×
×
= × = = ∑ ∑ ∑ = = =

A scheme was developed to select food items to represent the total diet
consumed in Taiwan. Based on the NAHSIT database, the total diet was classified
j
2

obtained by chemical analysis of the analytes in food samples, while CR and BW
values were derived from the database of the Nutrition and Health Survey in Taiwan
(NAHSIT). The HI values were calculated using the equation below and expressed as
% ADI.
100% ADI BW
CR C
100% ADI
ADD
%ADI HI
n
1 i P j
ij i
n
1 i P
ij
n
1 i
ij j
×
×
×
= × = = ∑ ∑ ∑ = = =

A scheme was developed to select food items to represent the total diet
consumed in Taiwan. Based on the NAHSIT database, the total diet was classified
A scheme was developed to select food items to represent the total diet consumed in Taiwan. Based on the NAHSIT database, the
total diet was classified into 12 categories and 47 sub-categories. By ranking food consumption rates, 128 food items were selected to
form a basic list for the TDS, which represented 83% of the total diet consumed daily in Taiwan. Taking into consideration the occurrence
pattern of the analytes in foods, the list was adjusted to a total of 97 food items for sample collection and preparation for chemical analysis.
Food products were purchased from 50 strategic sampling sites throughout Taiwan over two seasons (spring/summer and autumn/winter)
for two consecutive years. Food samples were prepared according to the conditions of serving and were homogenized and mixed prior to
chemical analysis to obtain data on C, of which the maximum concentration of each food item was used to make a conservative estimate.
The BW of nine age groups and CR of individual food items by each age group were used as exposure factors in risk calculation. The results
indicated that the levels of exposure to benzoates and sorbates of the general consumers at 95th
percentile in Taiwan were below 50% ADI
for benzoates and around 10% ADI for sorbates, which suggest relatively low risk of concern. The present TDS work has helped to renew
and strengthen the framework of TDS in Taiwan and facilitate further similar risk assessment work on other target chemicals in food.