Article Text


Infant care practices related to cot death in Turkish and Moroccan families in the Netherlands
  1. B E van Sleuwen1,
  2. M P L’Hoir1,
  3. A C Engelberts3,
  4. P Westers4,
  5. T W J Schulpen2
  1. 1Department of Medical Psychology of the Division of Paediatrics of the University Medical Centre, The Wilhelmina Children’s Hospital, Utrecht, Netherlands
  2. 2Centre for Migration and Child Health, Division of Paediatrics of the University Medical Centre, The Wilhelmina Children’s Hospital
  3. 3University Hospital Leiden, Department of Paediatrics, Leiden, Netherlands
  4. 4Centre for Biostatistics, University of Utrecht, Netherlands
  1. Correspondence to:
    Mw drs B E van Sleuwen, University Hospital Utrecht/Wilhelmina Children’s Hospital, KA.00.004.0, PO Box 85090, 3508 AB Utrecht, Netherlands;


From 1979 to 1993 Turkish infants had a significantly higher cot death risk compared to Dutch infants. In contrast Moroccan infants had a risk of cot death that was approximately three times lower compared to Dutch infants during the same period. This study shows that these differences have disappeared, while differences still exist in infant care practices between these ethnic groups. At 28 well-baby clinics, questionnaires were distributed for this sample selection. The response was 82%. Data were collected on 55 Turkish, 54 Moroccan, and 210 Dutch families.

Less than 7% of these three ethnic groups still placed infants in the prone position. Moroccan mothers hardly smoked. Turkish people used pillows and Moroccan people used soft mattresses more often. Moroccan families practised swaddling more widely. Length of maternal residence influenced some care giving practices. As a result of this study, subgroup specific intervention campaigns for safe sleeping can be developed for Turkish and Moroccan families.

  • cot death
  • sudden infant death syndrome
  • infant care practices
  • ethnic group

Statistics from

The Netherlands with its 16 million inhabitants has a relatively large ethnic minority of 1.3 million persons, amounting to nearly 9% of the population. The largest group came to our country as migrant labourer in the 1960s from Islamic Mediterranean countries, chiefly Turkey and Morocco. The other major group consists of people from Surinam in South America and from the Netherlands Antilles in the Caribbean, both former Dutch colonies. Nearly half of the migrant population lives in the four major cities (Amsterdam, Rotterdam, Utrecht, and The Hague), where 15% of the population and 40–50% of the children are of foreign descent.

In a retrospective analysis of 20.211 death certificates of all children who died between 1979 and 1993, data about cot death were collected for different ethnic groups.1,2 The incidence of sudden and unexpected death was 1.65 per 1000 live born Turkish infants. In contrast, the Moroccan infants had an incidence of 0.40. Compared to Dutch infants, whose cot death incidence was 1.4 per 1000 live births, the risk of cot death was significantly higher in Turkish infants and more than three times lower in Moroccan infants. There was no readily available explanation for this difference, except that Turkish fathers as well as mothers smoke far more heavily than Moroccan parents. To explore the reasons behind these incidence differences further, the current study on infant care practices was performed.

In 1999, Van der Wal and colleagues3 examined the care practices of 1529 infants—114 Turkish infants, 225 Moroccan infants, 181 Surinam infants, 775 Dutch infants, and 234 infants of other descent. Prenatal smoking was low in Moroccan mothers (0.4%), while Turkish mothers smoked almost as much as Dutch mothers (18.5% and 21.6% respectively). Daily smoking in the home occurred more often in Turkish families, compared to the Moroccan and Dutch homes. Turkish infants were placed prone more often than Moroccan and Dutch infants, but this difference was not statistically significant. Turkish and Moroccan families did not differ concerning breast feeding. The use of duvets and a pillow was higher among Turkish and Moroccan infants compared to Dutch infants.

Several risk and preventive factors for cot death were not covered in the study of Van der Wal and colleagues,3 however. In order to investigate what other differences between ethnic groups exist in infants care practices, the present study was carried out. Next to the previous mentioned aspects of infant care, the study focused on overheating, ventilation, room and bed sharing, type of mattress, dummy use, and swaddling.


A pilot study was carried out, including 22 Turkish, 44 Moroccan, and 31 Dutch families living in the Utrecht area.4 After analysis the questionnaire was adjusted on several items and prepared for the present study. From the Dutch Central Bureau of Statistics a list of villages and cities was gathered in which more than 500 Turkish and more than 500 Moroccan families lived.5 Eight regions, including urban and suburban areas, were randomly selected. It was a representative sample of concentrations of Turkish and Moroccan populations. Data were collected by face to face interviews from June 1999 to October 1999 by two previously trained interviewers at 28 well-baby clinics. From the pilot study we learned that many ethnic mothers who did not speak Dutch, were accompanied by a family member who did speak Dutch. For practical and financial reasons it was decided not to use interpreters. During consulting hours at well-baby clinics all parents of infants between 4 weeks and 2 years of age, not only Turkish and Moroccan and Dutch parents, were asked to participate. Ethnicity of the child was defined as the ethnicity of the mother. The final question in the questionnaire was if the parents had ever heard of cot death.

The thermal resistance of the total number of layers bedding was defined by tog values. A tog is a unit of thermal resistance, one tog being 10 times the temperature difference in degrees Celsius between the surfaces of a material when the heat flow through it is 1 watt/m2.6 Tog values were estimated from the usual use of bedclothes as reported by the parents.

Statistical methods

Correlation analyses, using Spearman’s rho and logistic regression analyses were used to analyse the data, using the Statistical Package for the Social Sciences (SPSS) version 8. χ2 tests, analysis of variance, and t tests were used to analyse the demographic data and the tog values. Confounders were selected in advance from earlier Dutch studies, literature review, and ongoing international research. They were: age of the mother, age of the infant, birth order, and maternal education. All p values are two tailed.


Data were collected on 55 Turkish, 54 Moroccan, and 210 Dutch families, and 33 families of other ethnic decent, giving a response of 83%. The 33 families of other descent were subsequently excluded in the study, as the number of families per ethnic group was too small. Forty one parents refused participation, four infants in retrospect did not meet the age criteria, and 32 parents did not participate because of language difficulties.

The following demographic data were collected: gender, birth order, birth interval, parental education, maternal smoking at time of the interview and maternal smoking during pregnancy, age of the infant, age of parents, age of mother at first live birth, country of birth of the parents, and length of stay in the Netherlands (table 1). The Turkish and Moroccan people did not differ in most of the demographic characteristics, except that the Moroccan infants were significantly younger than the Dutch and Turkish infants. The Dutch parents were higher educated. We checked by multivariate analysis whether some age related factors, such as room sharing, bed sharing, use of a pillow, use of a duvet during summer and winter, and dummy use were confounded. Some of these factors were influenced by age; for those factors, age specific groups were made.

Table 1

Demographic factors

Sleeping position

The prevalence of prone sleeping was very low (<10%) in all groups (table 2). Moroccan parents and Turkish parents placed infants in the side position to sleep more often, and in the supine position less often compared to Dutch parents.

Table 2

Relation between ethnicity and demographic factors* on sleeping position and bedding


Pillows were used more often in Moroccan families than in Dutch ones, but in Turkish families they were used even more so (table 2). Moroccan people compared to the Dutch families used soft mattresses most often. Plastic covers to protect the mattress are used equally often by all three groups. Cot buffers were used equally often in Turkish and Moroccan families, but significantly more than in Dutch families.

Fifty three per cent of the Turkish, 35% of the Moroccan, and 71% of the Dutch group used a Dutch sleeping sack. Blanket use, however, did not differ significantly between the groups, neither during summer nor during wintertime, and is therefore not noted in the table.

In this study only 25% of the Dutch parents used duvets. However, 55% of both Turkish and Moroccan people used duvets in wintertime.

Tog values

The Turkish and Moroccan population use duvets in summertime more often than Dutch families. However, after adjustment for “birth order” and “level of education of the mother”, the difference is not statistically significant. Low level of education was related to duvet use in summertime (r = 0.24, p < 0.0005). The estimated total amount of tog values in summer is significantly higher for Turkish (mean 4.71 (SD 3.95)) and Moroccan people (3.73 (SD 4.11)) compared to Dutch people (mean 2.55 (SD 3.03); p < 0.0005). In wintertime the statistically significant difference in tog values between these three groups remains (p = 0.005).

Central heating

In summer and in wintertime, Turkish and Moroccan families have the central heating on more often compared to Dutch families (table 3). The room temperature in one fourth of Turkish and Moroccan families is higher than 20 degrees, while this is the case in 17% of the Dutch families. These differences are not statistically significant, however.

Table 3

Relation between demographic factors* and ethnicity on parental room sharing, bed sharing, sleeping separately, and swaddling

Keeping a window open in the infant’s room seems to be culturally related; one third of the Dutch parents, less than one fourth of the Turkish, and one sixth of the Moroccan families have a window open. It was also asked whether the parents ventilated the infant’s room. Almost all parents of all ethnic groups either did ventilate the room or had a door or window open in the infant’s room.

Bed sharing, room sharing, and leaving the door open

Room sharing (excluding bed sharing) occurred in one third of the Moroccan families (table 3). When the child sleeps in a separate room, Moroccan families often keep the door of that room open (OR 3.13, 95% CI 1.15 to 8.33), compared to Dutch families.

Maternal smoking

Moroccan mothers smoked significant less during pregnancy and at the time of the interview than Turkish and Dutch mothers (χ2, p = 0.01 resp. χ2, p = 0.006).

Dummy use, thumb/finger sucking, and mouth breathing

Moroccan infants were offered a dummy most often. In the multivariate analysis, however, after introducing the factor “birth order”, the difference loses its statistical significance. Moroccan infants appear to breathe through their mouth* and suck their thumb/finger less often than Turkish and Dutch infants. The mouth breathing effect disappeared in the multivariate analysis but thumb/finger sucking remained statistically significant (p < 0.01). Thumb sucking occurs in 50.5% of the Dutch, in 40.0% of the Turkish, and 35.2% of the Moroccan infants. The correlation between dummy use and thumb/finger sucking is statistically significant (r = −0.12, p = 0.03) as well as between thumb/finger sucking and mouth breathing (r = 0.13, p = 0.03), although it is not very strong.


Moroccan infants were swaddled more often compared to Dutch and Turkish infants (OR 2.63, 95% CI 1.03 to 6.73) (table 3). In the pilot study we had already found that Moroccan people swaddled their infants more often (29.5%) than Dutch (0%) and Turkish (13.7%) families. In the current study the mothers themselves had been swaddled more often in the ethnic groups than the Dutch mothers had been (Turkish 86.3%, Moroccan 75%, and Dutch 6.7%).

Information about prevention of cot death

Half of the Turkish and Moroccan families did not know anything about cot death. Forty five per cent of the Turkish people and 52% of the Moroccans had never heard of it before. Only 3% of the Dutch people had never heard of cot death before.

Length of maternal residence in the Netherlands

A longer period of stay in the Netherlands is significantly (p < 0.05) correlated with less use of a pillow (r = 0.25), less use of a cot buffer (r = 0.26), and less use of central heating (r = 0.21). Length of maternal stay had no effect on smoking behaviour of the mothers.


A large study of infant mortality in immigrant groups, covering the period from 1979 to 1993, showed an increased risk of cot death for Turks, but a decreased risk for Moroccans compared to Dutch infants.1 However, the increased cot death incidence of the Turkish decreased proportionally with the Dutch decrease in incidence.8,9 All in all, cot death differences between the Turkish and Moroccan populations have almost disappeared (1996–2000, Turkish 0.24, Moroccan 0.28, and Dutch 0.16 per 1000 live born infants). This study looks at cultural differences in infant care practices related to cot death in the light of a historical difference in incidence between three ethnic populations. Understanding such differences is important in furthering knowledge of how infant care practice may relate to cot death.

A limitation of this study is a response of 83%, and the lack of data on the non-response group.

Summarising, both Turkish and Moroccan mothers were less likely to lay their baby supine, less likely to use a sleeping sack, and more likely to use a pillow or cot buffer. Turkish mothers in particular were more likely to use a pillow, while Moroccan mothers were more likely to have a soft mattress. Both had higher tog values both in the summer and winter. Conversely there were protective factors in that both groups were more likely to share the parental room without bed sharing, less likely to smoke (especially Moroccan mothers), and more likely to swaddle.

A full explanation for the rather constant incidence of cot death in Moroccan infants cannot be given. Moroccan people have several customs which are known to be associated with a decreased risk of cot death, such as little maternal smoking and room sharing, and perhaps factors such as swaddling and leaving the bedroom door open, thus improving parental surveillance. Since we know that half of the Dutch children that are still being placed prone (8%) are children that cry excessively,9,10 we speculated that in the Moroccan population the prone sleeping group could be underrepresented, because they swaddle these infants.11

Finally, it has been described that people from Southern European countries promote dummy use as a measure to prevent thumb sucking. Mouth breathing occurs less in these countries.7,12 It has been established in many studies that dummy use, especially during the last sleep, lowers the risk of cot death.12–16 It has been reported that people from Southern European countries (Turkey, Spain, and Morocco) have a low prevalence of thumb sucking because mothers consciously prevent this by giving the breast or a dummy.7 The social stigma against thumb sucking in the Southern European cultures thus appears potentially protective against cot death.

However, since most risk and protective factors are shared by both Turkish and Moroccan families, cultural infant care practices cannot really account for the observed differences.

Information on cot death does not seem to reach this group of young ethnic families. Many parents did not even know of the concept of cot death, let alone of important risk factors.

Differences that exist in infant care practices between Turkish and Moroccan populations have their origin in old habits and customs that by chance seem to be either preventive or risk increasing. It is very important for the prevention of cot death to keep in mind the ethnic differences in childcare practices. It is important to inform ethnic minorities about their native protective customs such as non-prone sleeping and perhaps swaddling which was recently postulated as being protective,17 and to warn them against a preference for side sleeping and pillows, and against several Western habits, such as the use of a duvet. The length of maternal residence did influence several risk and preventive factors positively, which suggests that the Dutch prevention campaigns reached these parents. Special attention in health education of ethnic groups should be given to overheating of the child, smoking, the use of a duvet, the number of layers used in bedding, and the use of a pillow. As a result of this study language specific intervention campaigns for safe sleeping can be developed for Turkish and Moroccan families.


View Abstract


  • * “Sleeping with an open mouth” is termed “mouth breathing”; however, only 50% of infants who sleep with an open mouth are in fact “mouth breathing”; the other 50% still breathe through the nose despite an open mouth, by pressing the tongue to the palate.7

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.