- Top of page
Objectives To investigate the association between iodine status and reproductive failure in a population of West African women.
Design Epidemiological survey on iodine deficiency disorders carried out in 1996–1997.
Setting The iodine deficient areas of Senegal (Casamance and Senegal Oriental).
Population Four thousand nine hundred and eighty women, aged 10 to 50, of whom 1544 adolescent and 462 pregnant women were examined for thyroid size and urinary iodine excretion. Their iodine status was associated with their fertility rate and reproductive failures.
Results Reproductive failure (defined as repeated miscarriages and stillbirth) was associated with low iodine status, with severe iodine deficiency increasing the risk. Poor nutritional status and illiteracy had a significant effect on the outcome of pregnancy: underweight women had a fourfold higher risk, and those who were illiterate an eightfold higher risk, of failed pregnancy, compared with nutritionally healthy, literature women.
Conclusions These findings emphasise the need to implement an effective iodine supplementation programme targeted at young and pregnant women in this area of western Africa. They also emphasise the importance of improving the nutritional status of young girls and the crucial role played by education in the prevention of reproductive failure.
- Top of page
According to the World Health Organisation more than one billion people in the world are affected by iodine deficiency because in numerous areas of the globe, and in Africa in particular, the soil and the water, and hence the locally grown food supply, have extremely low levels of iodine. Iodine deficiency manifests itself most obviously as a goitre in women, and as cretinism in young children, but it also has numerous consequences on reproduction: miscarriage; premature births; stillbirths; elevated infant mortality rates; and hypothy-roidism in the newborn. These health problems have been grouped under the heading iodine deficiency disorders1,2. Following the International Conference on Nutrition in Rome in 1992, the eradication of the iodine deficiency disorders is now a primary goal in numerous countries, particularly in Africa, where many governments have opted for the strategy of iodination of salt.
An epidemiological study was conducted during 1996 and 1997 in Senegal on the prevalence of the iodine deficiency disorders in two iodine deficient areas, Senegal Oriental and Casamance. More than 9000 men and women between the ages 10 and 50 years participated in this study. The data gathered from 4980 adolescents and adult females gave evidence of the relationship between iodine status and the frequency of repeated miscarriage and stillbirths in women.
- Top of page
The Republic of Senegal is situated on the Atlantic coast of Africa. The two regions studied are in the western and southern parts of the country. The climate is sub-Saharan, rainfall is rare and the soil has a very low content of iodine. The majority of the 1.9 million inhabitants living in the regions are dispersed among villages established around sources of water where the level of iodine is very low. They consume essentially locally produced food (millet, sorghum, groundnuts, and cassava), and the level of iodine in the diet is very low.
Cluster sampling randomly chose the women studied. They underwent clinical examination of the thyroid gland and anthropometric measurements. Urine samples were collected, the iodine level measured, and they were questioned by health personnel about their menstrual cycles and their obstetric histories, including the numbers of miscarriages and stillbirths.
The data were analysed using the statistical software Epi Info 5.
- Top of page
Four thousand nine hundred and eighty women participated in this study of whom 1544 (31%) were adolescents between the ages of 10 and 18 years and 462 (9%) were pregnant women (Table 1).
Table 1. Composition of the sample. Values are given as n or n (%).
|Age group (years)||No. of women||No. of goitres|
Of the 4980 women, 32% showed evidence of a goitre, where the average rate of occurrence in the general population is 20%. All the goitres found were of average size (Stage 1 a and 1 b of the De Maeyer classification). The WHO considers West Africa to be a region were iodine deficiency is severe, since the prevalence of goitre is above 30%3. In our study the prevalence was virtually identical among adolescent and adult women.
The World Health Organisation4 recommends that the severity of iodine deficiency is determined in a population by measuring the level of iodine excreted in the urine using the Dunn technique5, on a sufficient number of samples. The iodine status of the population is then estimated from the median level of urinary iodine (expressed in μg/L urine). If the median is less than 20 μg/L, the deficiency is considered as severe; if the median is between 20 and 50 μg/L, the deficiency is considered as moderate and if the median lies between 50 and 100 μg/L it is considered as slight.
According to the classification of the World Health Organisation detailed above, the population is considered as ‘moderately’ iodine deficient with median levels of urinary iodine between 20 and 50 μg/L. The urinary iodine levels were virtually identical in adolescent and young adult women but were significantly higher among women over 30 years old (Table 2).
Table 2. Table 2. Distribution of urinary iodine of nonpregnant women according to age. Values are given as n, unless otherwise indicated. SD = standard deviation.
|Age (years)||No. of women||Median (pgL)||Mean (μg/L)||SD (μg/L)||25th centile (μg/L)||75th centile (pg/L)|
|TOTAL||4498||43||60|| || || |
Four hundred and sixty-two women (9%) stated that they were pregnant. Each of these women was compared, as far as possible, with a nonpregnant woman of the same age, residing in the same village and, therefore, having access to the same sources of alimentary iodine (Table 3). Goitres were significantly more frequent in pregnant women (pregnant 42%; non-pregnant 34%; χ2= 5.88; P < 0.05). Moreover, pregnant women tended to excrete larger quantities of iodine in the urine (t= 2.70; P < 0.01).
Table 3. Comparison of iodine excretion between the group of pregnant women and the group of nonpregnant women. Values are given as n or n%. SD = standard deviation.
| ||No. of women||No. of goitres||Urinary iodine Mean (μ/L)||SD (μg/L)|
|Pregnant women||462||194 (42)||60||39|
|Nonpregnant women||420||143 (34)*||52||41†|
Iodine deficiency and presence of a goitre
Of the 2954 women over the age of puberty (Table 2) the reproductive history was known in 2494 (84%). These women had 11,223 live births and 715 reproductive failures, where reproductive failure is defined as a history of at least 3 miscarriages or one or more stillbirths. The infant mortality rate of the 11,223 live births was not determined during the study. Table 4 shows a dose-response relationship between iodine deficiency and reproductive failure. For iodine deficiency overall the odds ratio (95% CI) of reproductive failure is 2.02 (1.68, 2.43). The presence of a goitre was not related to the frequency of reproductive failure, as illustrated in Table 5. For women with a goitre, the odds ratio (95% CI) of a failed pregnancy was 1.17 (0.98, 1.40).
Table 4. Number of women with reproductive failure and iodine status. Values are given as n or n (%), unless otherwise indicated. OR = odds ratio; WHO = World Health Organisation.
|Mean urinary iodine (μg/L)||Iodine deficiency (WHO definition)||No. of women||No. of miscarriages ≥= 3||No. of stillbirths||Total reproductive failures||OR (95%CI)|
|<20||severe||566||238||38||276 (49)||3.64 (2.92, 4.55)|
|20–50||moderate||561||121||31||152 (27)||1.42 (1.12, 1.80)|
|51–100||mild||305||53||14||67 (22)||1.08(0.79, 1.47)|
|101–150||none||1062||176||44||220 (21)||1 (reference)|
|TOTAL|| ||2494||588||127||715 (29)|| |
Table 5. Number of women with reproductive failure and presence of a goitre. Values are given asnorn (%).
| ||No. of women||No. of women with reproductive failure|
|Goitre present||897||276 (31)|
|Goitre absent||1597||439 (27)|
Given the prevalence of malnutrition in this region where approximately 25% of adult women are underweight, an attempt was made to determine a relation between reproductive failure and malnutrition. Body mass index was calculated for 2421 women sampled over the age of 17. They were classified as malnourished (body mass index ≤ 18.5) or well nourished (body mass index ≥ 18.6), and their nutritional status was associated with their pregnancy outcome (Table 6). A significant difference (P < 0.0001) was found between underweight women and women who were not underweight. The underweight women had an odds ratio (95% CI) of a failed pregnancy of 1.86 (1.51, 2.28) compared with women in good nutritional status.
Table 6. Number of women with reproductive failure according to body mass index (BMI). Values are given as n or n (%).
| ||Reproductive failure||No reproductive failure||TOTAL|
|BMI ≤ 18.5||210 (35)||395 (65)||605|
|BMI ≥ 18.5||404 (22)||1412 (78)||1816|
|TOTAL||614 (25)||1807 (75)||2421|
A stratified analysis was conducted to determine whether nutritional status would have a modifying effect in the relation between iodine deficiency and reproductive failure (Table 7). The analysis shows that nutritional status has a modifying effect on the relationship between iodine status and reproductive failure: underweight women who also had iodine deficiency had a higher risk of having reproductive failure compared with women of normal weight.
Table 7. Effect of nutritional status on reproductive failure according to iodine status. Values are given as n, unless otherwise indicated. BMI = body mass index.
|Nutritional status||Iodine status (μg/L)||Reproductive failure||No reproductive failure||Odds ratio (95% CI)|
|Underweight||< 50||145||129||4.60 (3.16, 6.71)|
| (BMI ≤ 18.5)||≤50||65||266|| |
|Normal weight||< 50||201||620||1.26 (1.01, 1.59)|
| (BMI ≤ 18.5)||≤ 50||203||792|| |
|TOTAL|| ||614||1807|| |
In the same manner, the degree of literacy of the women was associated with their reproductive failure (Table 8). Among the 2490 women, 62% were illiterate. A significant difference (P < 0.0001) was found between literate and illiterate women: 41% of illiterate women experienced reproductive failure, compared with 8.5% of the literate women. Compared with literate women, illiterate women had an odds ratio (95% CI) of reproductive failure of 7.53 (5.82, 9.73).
Table 8. Number of women with reproductive failure according to degree of literacy. Values are given as n or n (%).
|Degree of literacy||Reproductive failure||No reproductive failure||TOTAL|
|Illiterate||634 (41)||907 (80)||1541|
|Literate||81 (8.5)||867 (91.5)||948|
|TOTAL||715 (29)||1774 (71)||2489|
A stratified analysis was conducted in order to assess whether the degree of literacy had a modifying effect on the relationship between iodine deficiency and reproductive failure (Table 9). In illiterate women the odds ratio (95% CI) of reproductive failure with iodine deficiency was 8.01 (6.31, 10.81), while in literate women the odds ratio (95% CI) of reproductive failure with iodine deficiency was 2.25 (1.37, 3.21).
Table 9. Effect of degree of literacy on reproductive failure according to iodine status. Values are given as n, unless otherwise indicated. OR = odds ratio.
|Degree of literacy||Iodine status (μg/L)||Reproductive failure||No reproductive failure||OR (95% CI)|
|Illiterate||< 50||480||254||8.01 (6.31, 10.18)|
| ||≥50||154||653|| |
|Literate||< 50||48||340||2.25 (1.37, 3.21)|
| ||≥ 50||33||527|| |
|TOTAL|| ||715||1774|| |
- Top of page
This study, as well as in other reports, shows that iodine deficiency does not seem to influence the age at menarche: 14.7 years in iodine deficient adolescents, compared with 14.6 years in noniodine deficient adolescents. A delay in the age of puberty has been described in another rural area in Senegal and has been attributed to malnutrition at an early age6
Several experimental studies in various animal species have shown that iodine deficiency has a dramatic effect on fertility. In sheep during pregnancy, for example, food deficient in iodine increased the number of abortions and stillbirths. Twenty-one percent of iodine deficient ewes had reproductive failure, compared with 4.3% of the controls7. The effect of iodine deficiency on the growth and development of farm animals living in an iodine deficient environment has been known for years8,9.
In women living in iodine deficient areas such as Papua New Guinea10, Central Africa11, and the Himalayas12, several epidemiological surveys have shown an increased risk of spontaneous abortions, stillbirths and neonatal deaths. Our study showed that women who have iodine deficiency have twice the risk of reproductive failure. Other nutritional factors such as selenium deficiency could also be responsible for reproductive failure. Nonetheless our data showed that the frequency of reproductive failure was directly proportional to the severity of the iodine deficiency (Table 4). In severely iodine deficient women (with urinary excretion of iodine below 50 μg/L) the rate of stillbirth was 15/1000 live births, whereas in nondeficient women (with urinary excretion above 50 μg/L) this rate was 8.7/1000 live births. One-third of iodine deficient women had more than three miscarriages, compared with 16% in nondeficient women. In Nepal, where iodine deficiency is severe and the prevalence of goitre is very high, the neonatal mortality associated with iodine deficiency has been estimated to be 16/1000 live births13, with a high proportion of anomalies of the central nervous system. In a severely iodine deficient area of Zaire, a survey has shown11 an association between reproductive failure and cretinism. We were unable to assess the prevalence of cretinism in our population.
The influence of iodine deficiency on reproductive failure is clearly demonstrated when iodine is administered to women before pregnancy. Prenatal iodine supplementation has been tried in New Guinea10, which was associated with a reduction in reproductive failure from 24% to 16%. Similar programmes in Tasmania were associated with a reduction in the rate of stillbirth from 20/1000 to 12/100014, and in Zaire from 18/1000 to 9/100011.
There are major changes in maternal thyroid hormone metabolism during pregnancy15–17. Up to eight weeks of gestation there is a rapid rise in levels of thyroxine-binding globulin followed by a more gradual rise. Rising oestrogen levels stimulate the synthesis of thyroxine-binding globuline, resulting in an increased output of thyroxine by the thyroid to maintain a normal free thyroxin level. If iodine is in limited supply, adjustment to these changes may be impaired leading to hypothyroxinaemia, elevation of serum thyroid stimulating hormone and increased thyroid volume18,19. In addition, there is an increased excretion of iodine in the urine during pregnancy, as we have found in this study (Table 3). The deficiency of maternal thyroxine results in decreased availability of thyroxine for early development of the fetus prior to the formation of the fetal thyroid gland. Maternal hypothyroidism seems consistently to increase the risk of miscarriage and stillbirth20,21. In iodine deficient areas22 maternal hypothyroidism, fetal hypothyroidism and fetal deprivation of elemental iodine may all occur to some degree. Neonatal hypothyroidism has been detected in 4% of newborn infants in India13 and in 10% of newborn infants in Zaire1.
In countries where the natural iodine supply is abundant it has been estimated that spontaneous miscarriage is twice as frequent in nontreated hypothyroid women as in euthyroid women, and the frequency of congenital abnormalities is increased by 20%. Perinatal mortality is also high and half of the children born to hypothyroid mothers have some degree of psychomotor abnormality23,24. The frequency of those abnormalities decreases dramatically with proper treatment of the thyroid insufficiency.
Our survey shows a very significant relationship between poor nutritional status in the mother and reproductive failure, the risk being nearly doubled (Table 6). It has been known for some time that during pregnancy chronic maternal energy deficiency may cause fetal malnutrition, low birthweight and increased perinatal mortality25. In England in the 1930s the neonatal mortality rate was 60–70 per 1000 pregnancies; by 1945 this had reduced to 40–50 per 1000 pregnancies. This improvement has been associated with efficient food distribution and vitamin supplementation of pregnant women during the Second World War26.
In our study the association of illiteracy and reproductive failure was striking, where 41% of illiterate women suffered reproductive failure, compared with 9% of literate women. Stratified analysis showed that illiterate women were eight times more likely to experience reproductive failure if they were deficient in iodine, compared with twice as likely if the women were literate. According to World Bank estimates primary school education reduces infant mortality by 36/1000, secondary school education by 72/1000 and that each additional educational year of the mother decreases infant mortality rate by 3.5% on average27. Teaching young women to read and write also delays the age at first marriage and increases the use of contraception. Contraception was used by 1% of the illiterate women, 6% of women with primary education and 22% of women with secondary education. In Senegal the percentage of girls in full-time education has increased regularly in the last decades but this survey was made mainly in remote rural areas where school attendance is lower than in other parts of the country27.