Urinary iodine excretion is a good marker of the very recent dietary intake of iodine and, therefore, is the index of choice for evaluating the degree of iodine deficiency and of its correction. Iodine concentrations in casual urine specimens of children or adults provide an adequate assessment of a population iodine nutrition, provided a sufficient number of specimens is collected. Twenty four hours samples are difficult to obtain and are not necessary. Relating urinary iodine to creatinine is expensive and unnecessary (2).

Several methods of determination of urinary iodine have been reported (124 - 127), including a fast colorimetric method allowing one technician to easily measure 200 samples in a working day (128). The most commonly used is called method A (2). Small samples of urine are digested with ammonium persulfate at 90-110°. The iodine content is then determined with the sensitive colorimetry of the Sandell-Kolthoff reaction (129) in which iodine is determined from its catalytic reduction of cerium ammonium sulfate in the presence of arsenic acid. This method emphasises urinary iodine concentration in the range 0-100 μg/L (0-1.19 μmol/L).

For epidemiological studies, the population distribution of urinary iodine is required rather than individual levels. Because the frequency distribution of urinary iodine is usually skewed towards elevated values, the median is considered instead of the mean as indicating the status of iodine nutrition. Table 5 shows the epidemiological criteria presently recommended for assessing iodine nutrition based on median urinary iodine concentrations in school-age children. Distinction has been introduced between iodine intake and the status of iodine nutrition of a population.

There is no need for the more difficult 24 hour urine collections. Casual samples from school children can be collected at the same time as the goiter is assessed.

In the past, levels have often been expressed per gram of creatinine excretion (130, 131). More recent studies (132) have indicated that the creatinine level is variable depending on the general nutritional status of the population. This contributes an independent source of variation which invalidates the ratio. Urinary creatinine also decomposes after three days without refrigeration, whereas urine iodine remains stable for months. The values of urinary iodine can be most conveniently expressed as a range with a median or by the proportions at a series of cut off points, < 20 ug /l, < 50 ug /l, and < 100 ug /l (2).

The size of the thyroid gland changes inversely in response to alterations in iodine intake, with a lag interval that varies from a few months to several years. The prevalence of goiter is an index the degree of longstanding iodine deficiency and, therefore, is less sensitive than urinary iodine in the evaluation of a recent change in the status of iodine nutrition (2).

Thyroid size is traditionally determined by inspection and palpation but ultrasonography of the thyroid provides a more precise and objective method.

The former statement that « a thyroid gland whose lobes have a volume greater than the terminal phalanx of the thumb of the person examined will be considered goitrous » remains valid (121). Table 6 shows the revised and simplified classification of goiter. Figure 6 shows large multinodular goiters stage II.

However, the evaluation of the prevalence of goiter based on palpation has been questioned because the reproducibility of assessment by palpation is low, especially with the size estimation of smaller glands, particularly in children (4).

Therefore, the method of choice is now ultrasonography which is reproducible with a maximum deviation of 10 %. Normative values for thyroid volume measured by ultrasonography as a function of age, sex and body surface area have been proposed by Delange et al. (133). However, these normative values might have been over-valued by some 30 % due to interobserver variability in thyroid ultrasonography (134). Updated normative values have been proposed on the basis of data collected in different continents in areas which never experienced iodine deficiency in the past. (135). By definition, a thyroid is considered as goitrous when its volume is above the percentile 97 established for sex, age and body surface area in iodine replete populations (133, 135).

The prevalence of goiter in iodine replete populations is below 5 % (4).

The serum thyroid hormone levels are a further index of the effects of iodine deficiency.

However, difficulties are often encountered in obtaining venous blood samples in populations due to apprehension about blood collection and operational difficulties. Therefore, these measurements are not routinely recommended in routine assessment and monitoring.

In spite of the difficulties in blood collection, it has to be kept in mind that the final objective of correction of iodine deficiency is not only to increase the access of the population to iodized salt and to normalize the urinary iodine concentration but mostly to normalize thyroid function tests (136).

Elevated serum TSH, unless exceptional pathological situations, indicates an insufficiency in the saturation of the T3 receptor in the brain, whatever the level of serum thyroid hormones. Therefore, elevated serum TSH constitutes an indicator of the potential risk of iodine deficiency on brain development.

Serum T4 and T3 are less specific indicators of iodine deficiency because they are modified usually only in conditions of at least moderate iodine deficiency (2). Moreover, these levels are largely influenced by age and sex (63). In moderate and severe iodine deficiency, serum T4 is low but T3 is variable, occasionally, high due to preferential T3 secretion by the thyroid. Elevated serum T3 in spite of low serum T4 is considered as a protective mechanism to most parts of the body, except the brain, where T3 is produced locally and not derived from the circulating T3 (137). Serum thyroglobulin represents a sensitive index of a state of thyroid hyperstimulation (104, 138).

A biochemical picture associating elevated serum TSH in spite of normal serum T4 and T3 is called subclinical hypothyroidism while overt hypothyroidism associates elevated TSH and low T4 with variable levels of T3.

The use of whole blood from finger pricks spotted on filter paper cards can be used at least for the measurement of serum TSH and serum thyroglobulin used as indicators of thyroid hyperstimulation and the consequence of the state of hyperstimulation respectively (139, 140,). A frequency distribution of serum TSH in neonates shifted to high values is a particularly sensitive index of the risk of potential damage of the developing brain due to iodine deficiency (139). In normal conditions, less than 3 % of neonatal TSH are above the critical threshold of 5 mU/L whole blood (4, 140). The validity of this threshold has been recently confirmed in Switzerland (140 bis) where, following an increase in the level of salt iodization from 15 to 20 mg/kg, the median urinary iodine of pregnant women increased from 138 μg/l, a borderline low value, to a normal value of 249 μg/l, while the frequency of neonatal TSH above 5 mU/L decreased from 2.9% to 1.7%. However, because of technical and financial limitations (141, 142), this variable is only rarely used as monitoring tool (143, 144) and has been recommended only in countries and areas where a program of systematic neonatal hypothyroid screening is already implemented (2, 4).