Autors

Gabriela Comisel, PhD, family practitioner, Craiova, Romania 

Frank Buntinx, MD, PhD., Professor, Department of General Practice, University of Maastricht, The Netherlands and University of Leuven, Belgium

Abstract

Objective. The study evaluates the relation between the level of psychological stress during pregnancy and the emergence of obstetrical pathology.

Methods. This cohort study was realised in the Dolj district (Romania). 290 consecutive pregnant women were included from 15 randomly selected primary care medical offices (130 pregnancies) and from the obstetrical section of the largest hospital from Dolj (160 pregnancies), during a period of six months. For each subject we collected personal, clinical and laboratory data. At the first meeting with the doctor, the pregnant women filled in a stress questionnaire identifying 22 important psychologically stressing life events. The presence of obstetrical pathology divided pregnancies into two groups: with or without obstetrical pathology. The stress score divided the women into three groups: with a low, moderate and high level of stress. The categorised stress-score was related to the pregnancy outcome. We also studied the influence of other co-variables: age, level of education, place of residence, medical history and general clinical status.

Results. Compared with pregnancies with low stress, the relative risk of obstetrical pathology in pregnancies with a moderate level of stress was 1.09 (95% CI= 0.80-1.48) and in high stress pregnancies 1.45 (95% CI= 1.08-1.94).

Conclusion. Psychological stress is a significant risk factor for obstetrical pathology, although stronger predictors can overrule its influence, such as a history of previous pathology or previous abortions. The family doctor and the obstetrician should identify possible stressing factors and search for appropriate counselling and treatment strategies for pregnant women with a high level of stress.

Key words:psychological stress, obstetrical pathology, pregnancy, family doctor.

Introduction

 In Romania, since 1992 the statistics show a negative natural movement of population, and the process had emphasized during recent years (1). This process resulted from a number of evolutions:

The birth rate decreased from 16 live births on 1000 inhabitants in 1989 to 10,5 in 1998.Infant mortality below the first year of age increased from 21,2 deaths per 1000 live births in 1989 to 26,9 in 1998. The number of pregnancies decreased from 215.998 in 1990 to 189.690 in 1998.General mortality increased from 10,7 deaths in 1000 inhabitants in 1989 to 12 in 1998 (1, 2). We believe that a high level of psychological stress can affect these events.

Psychological stress appears when there is a discrepancy between the resources, abilities and individual capacities of a subject and his needs. In such conditions the psychological homeostasis is disturbed, which affects the somatic homeostasis (4). Stress inhibits the production of gonadotropin hormones and thyreostimulant hormones (5,6). These perturbations can produce obstetrical pathology. As a first step in testing this hypothesis, we prospectively examined if there is a relation between the level of psychological stress and the appearance of obstetrical pathology in pregnant women in the Dolj district (Romania).

Method

The design of the study

We performed a cohort study, with patients included from the 1^stof August 1998 till the 31^stof January 1999, in Dolj district, Romania and then followed up until delivery. We included pregnant women admitted in the obstetric section of the largest hospital from Dolj as well as cared for in primary care medical offices.

In the hospital’s obstetric section only abnormal pregnancies were admitted. These women originated from villages and towns of the Dolj district. In our study we included all pregnancies consecutively admitted to this hospital during the study period as far as the women consented.

To select the primary medical care offices we used stratified randomisation: we selected one medical office from a big village, and 14 offices from Craiova (both from downtown and sub-urban, and from the rich and the poor parts of the town). In this way we included women with different levels of income, education and access to medical care, with or without pathology. From these medical offices, 20 family doctors accepted to collaborate in this study. We consecutively included all pregnancies presented to these GPs during the study period.

The power of the study

 According to calculations made before the study, we required 65 subjects in each group (with the three different levels of stress) of pregnancies to be able to significantly detect a relative risk (RR) of 1,5, using a confidence level of 95%, a power of 80%, a rate of diseased in the non-exposed group of 50%, and an even number of women in all three groups.

Data collection

We collected data of all pregnancies admitted in the hospital twice a week. The researcher presented the stress test to each newly admitted patient. Once a week we visited each GP office and collected the data from the medical records. In primary care, the pregnant woman filled up the stress test at the medical office.

Each pregnant woman was admitted in the study only once. She filled in the stress test during the first contact with the doctor after the study had begun, independently of the foetal age at that moment.

The stress test (7) consists of 22 questions about the most important psychologically stressing life events. This questionnaire is a part of The Stress Questionnaire of Dr. Thomas Holmes from the Medical Centre of Washington University (8,9). This test is concise and provides both the level of stress and the predisposition for psychosomatic diseases. Each question has a dichotomised answer (Yes/No) that is quantitatively weighted. The level of stress is given by the sum of all coefficients of the Yes answers. The level of stress is considered to be low if the total coefficient is fewer than 30, moderate between 30 and 60 and high above 60. On a sub sample of 10 patients, we examined test-retest agreement per item with an interval of two days. Overall we found an accuracy of 0.92 (95%CI = 0.87-0.95) and a kappa value of 0.81 (95%CI = 0.73-0.89). According to Altman (10) this is a very good result.

Additionally, we collected information about the age, level of education, place of residence, medical history, and general clinical status of the pregnant woman.We followed, from records, every pregnant woman till she gave birth to see if any obstetrical pathology emerged during the pregnancy.

*Level of education classified the women into those with a low (8 classes or less) and a high level (more than 8 classes) of formal education.

*Age was categorised in two groups: between 16-35 years and less than 16 or more than 35 years.

*Age of the foetus (pregnancy stage) at the moment of the stress testing, split the subjects into three groups, according to the trimesters: first (0-12 weeks), second (13-28 weeks) and third trimester (29-42 weeks).

*Residence was classified as rural or urban (indicating a town with more than 20.000 inhabitants).

*Profession of the woman was categorised as without a job or any worker.

*Number of prior births classified the pregnancies in three groups: 1) without any previous birth, 2) 1-3 births, 3) more than 4 births.

*Number of prior abortions divided the pregnancies into a group without any abortion, with one or more abortions without pathology (at demand) and with at least one pathological abortion.

*Medical history identified women without any prior disease, with a previous obstetrical disease (threatened abortion, malformation of the baby, death of an infant of less than one year, premature baby, caesarean, sterility, abortion disease, rhesus incompatibility, ectopic pregnancy) and with a medical or gynaecological disorder (fibroids, pelvic infection, ovary cyst, breast cancer).

*Presence of pathology during the pregnancy stratified the subjects into two groups: normal (the pregnant women was in good health, she only came to a doctor for routine control) and pathological (the pregnant women had obstetrical problems during the pregnancy: threatened abortion, imminent abortion, intrauterine death of the baby, postmaturity pregnancy, hyperemesis or severe nausea, pre-eclampsia or eclampsia, incompetent cervix uteri, placenta praevia, ectopic pregnancy.

Analysis

The relation between the level of stress and the obstetrical pathology during the pregnancy was expressed by the relative risk (RR) of pathology and its 95% confidence interval (95% CI) in patients with a moderate and high level of stress compared with a low level. These data have been analysed with Epi-info 6.04.

Additionally we fitted a logistic model with the pregnancy outcome (pathological versus normal) as the dependent variable. The result of the stress test (3 classes) was the main independent variable and co-variables were age, residence (rural versus urban), level of education, professional group, number of prior deliveries and presence of prior abortions and previous medical history.

We entered co-variables progressively starting with background variables and indicators of socio-economic status and continuing with information on previous medical and obstetrical problems. SPSS was used for the regression analysis.

 Results

I. Description of the study group

              Before the start of the study one family doctor refused to participate in the study (meaning that three or four pregnancies were not included). After the beginning of the study two women from medical offices and three from the hospital refused to fill in the stress test.

            290 women were included in the study: 160 from the hospital and 130 from the primary medical care offices. 99% of the subjects were between 16- 35 years old .101 patients (35%) live in urban areas, 189 (65%) in rural areas. 6 (16%) patients had eight classes or less and 244 (84%) had more than eight classes of formal education. There was a significant relation between a low level of education and living in rural areas (RR = 2.57, 95% CI =1.96-3.38). 105 (36%) were unemployed, 185 (64%) were workers. 22% of the pregnant women entered in the study during the first, 62% during the second, and 16% during the third trimester. 223 (77%) of the patients were primi-gravidae, 22% had between one and three deliveries before this pregnancy and three (1%) patients already had four or more deliveries. There were 119 (41%) patients with at least one abortion in their history. Most patients (n=167, 58%) didn't have any known non-obstetrical disease during the pregnancy, 126 (43%) had a general disorder and 44 (15%) had a disorder of the genital tract.

The results of the stress test were summarised as high in 100 (34%), moderate in 133 (46%) and low in 57 (20%) pregnancies.  170 (59%) of the women had health problems during their pregnancy. These included obstetrical pathology in 64 (22%) subjects (threatened abortion, imminent abortion, intrauterine death of the baby, postmaturity pregnancy, hyperemesis and sever nausea, pre-eclampsia and eclampsia, incompetent cervix uteri, placenta praevia, ectopic pregnancy, ovary cyst, incontinence of the cervix uteri)and medical pathology in 59 (20%)subjects (endocrine, eye, digestive, cardio-vascular, rheumatologic, neurological, respiratory and surgical disorders or a tumour),

II. Relation between basic characteristics and the obstetrical pathology

Urban residence, low level of education, absence of a job, belonging to an extreme age group, prior deliveries or prior abortions and a positive medical history, all increased the likelihood of obstetrical pathology at bivariate analysis (table 1).

III. Relationship between obstetrical pathology and level of stress during pregnancy

Compared with pregnancies with low stress (<30), the relative risk of the obstetrical pathology is 1.09 (CI 95% 0.80-1.48; p=0.002) in the group with a moderate level of stress (30-60); and 1.45 (95% CI 1.08 –1.94; p=0.002) in the group with a high level of stress (>60)(table 1). The relative risk of obstetrical pathology in patients with a high level of stress (>60) is 1.33 (95% CI 1.09-1.63; p=0.014) compared with a moderate level of stress (<60).

During logistic regression analysis, the odds ratio of moderate stress compared to low stress remained statistically non-significant. The odds ratio of high stress was 2.15 (95%CI = 1.02-4.55) after adjusting for background variables (age-group, place of residence, number of previous deliveries) and indicators of socio-economical status (education, profession). Statistical significance was removed, however, if a history of previous abortions or the presence of pathology in the medical history was added to the model. If all co-variables were concurrently added to the model, the resulting odds ratios for the level of stress were 0.96 (95%CI = 0.43-2.14) for moderate stress and 1.09 (95%CI = 0.44-2.67) for high stress. The results of both models are presented in table 2. The two models had accuracy values of 59% and 80%. The Hosmer and Lemeshow test resulted in p-values of 0.24 and 0.47 respectively, indicating good fit.

IV. Causes of psychological stress during pregnancy

The stress test also evidenced the most frequent events associated with high level of stress (table 3). At the end of the test we directly asked the pregnant woman herself about the perceived causes of her stress (table 4). The most frequent causes of stress that were identified were the presence of the pregnancy itself, financial problems, difficulties in family relations and health problems. The results of the stress test are largely similar with the straightforward declarations of the women.

 
Discussion

Our bivariate results show an increase of the incidence of obstetrical problems related to the level of stress (measured by our stress test). This increase has a dose-response effect and is robust after adjusting for age, residence, level of education, professional group and number of prior deliveries as co-variables.

Both the magnitude of the odds ratio and the statistical significance disappear in the multivariate model after adding a history of previous abortions or a pathological medical history to the model. The latter are medical factors that can easily be related to the likelihood of obstetrical problems during a next pregnancy. Both variables are also significantly correlated with the level of stress (Spearman correlation coefficients: 0.16 and 0.26). These results indicate that increased risk resulting from previous or actual medical or obstetrical problems tends to overrule the impact of high stress, which is not unexpected.

The literature suggests that stress is stronger in the first and the third trimester of a pregnancy (11), but our data showed the highest levels of psychological stress in the first and the second trimester.

It is known that stress can cause psychosomatic disease (12,13,14,15,16,17), but it wasn’t clear if it also would influence the evolution of the pregnancy. This study is important because it shows a way to evaluate the level of psychological stress and its relation with the occurrence of obstetrical pathology. Our results support our hypothesis that psychological stress negatively influences the outcome of the reproductive process.

It has been mentioned before that people don’t know exactly the causes of their stress while an important step in helping to cope is to discuss exactly these reasons of stress. This study showed us that at least these pregnant women know pretty well their stressing problems, although they weren’t able to cope with them.

Two factors can be implicated in decreasing the level of stress: medical care and society. The family doctor and the obstetrician should identify pregnancies with a high level of stress and use medical and psychological therapy and counselling. Society, however, must improve the standards of life (assurance of an adequate home, income and accessibility to medical care) and so decrease the most frequent reasons of stress in pregnancy.

The stress test (7) used by us is a part of The Stress Questionnaire of T. Holmes of the Medical Centre of Washington University (8). In Romania it is the most frequently used stress questionnaire (8,9). It is possible that the use of another questionnaire could influence the magnitude of our results, but we have no reason to think that the presence or the direction of the relation between psychological stress and the outcome of pregnancy would be different. The results of the test- retest analysis showed a good reliability of this stress test.

Summarising, the results of this study support our hypothesis of a positive relation between stress during pregnancy and the occurrence of obstetrical pathology, although the stronger predictors, such as a history of previous pathology or previous abortions, can overrule its influence.

Acknowledgements

This research was performed within the framework of the Dutch-Romanian MATRA-program in which Romanian GPs receive extensive research training.

The authors acknowledge the help of Prof. Dr. GJ Dinant of the University of Maastricht, of the members of the MATRA research group, of the physicians who were helpful in collecting the data and of Berna Schouten (UM).

The theme of this study, the use of stress test was recommended by Prof. Dr. Al. Olaru of the University of Craiova. We thank him for his help. 

References

 1. Anuar of the Medical Statistic 1998. Ministerul Sanatatii, Centrul de calcul, Statistica sanitara si documentare medicala. Bucuresti;1997: XI-XXVIII :14, 56, 134.

2. World Health Organisation(US). From Contraception to reproductive health care. CEU Press;1997: 43-70

3. Honkanen H, Benagiano G. Reproductive health in Eastern Europe: an overview and the challenges ahead. In: Lindmark G, Horga M, Campana A, Kasone J, editors. Towards Better Reproductive Health in Eastern Europe. CEU Press, WHO; 1999: 1-18.

4. Lindmark G. Can a process of quality assurance improve pregnancy outcome? In: Lindmark G, Horga M, Campana A, Kasone J, editors. Towards Better Reproductive Health in Eastern Europe. CEU Press, WHO; 1999: 19-46.

5. Coculescu M. Neuroendocrinologie clinica. Bucuresti: Ed. Stiintifica si Enciclopedica; 1986: 223-233

6. Daly  MJ. Principles of Psychosomatic Medicine. In: John J., editors. Gynaecology and Obstetrics. Sciarra; 1989:6;70:1-7

7. Kunz JRM, editor. The effects of stress.In: Family Medical Guide: The healthy body. The American Medical Association. New York;1982:22.

8. Iamandescu I B. Psihologie medicala. Bucuresti: Infomedica; 1997:43-56.

9. Derevenco P, Anghel I, Baban A. Stresul in sanatate si boala. De la teorie la practica. Cluj-Napoca: Ed. Dacia; 1992:18-110

10. Altman DG. Practical statistics for medical research. London, Chapman & Hall; 1991.

11. Da-Costa D, Larouche J, Dritsa M, Brender W. Variation in stress levels over the course of pregnancy: factors associated with elevated hassles, state anxiety and pregnancy- specific stress. J Psychosom Res 1999 Dec; 47(6):609-21.

12. Valeanu V, Daniel C. Psihosomatica feminina. Bucuresti: Ed. Medicala; 1977.

13. Mackay B. The Stress and Anxieties of Pregnancy. In: Obstetrics and Newborn. 2nd rev ed. Sydney: W.B. Saunders Company; 1986: 68, 69, 86-90, 255- 259.

14. Restian A. Bazele medicinei familiei. Bucuresti: Ed. Medicala; 2000: 217-241.

15. Iamandescu I B.Stresul psihic si bolile interne. Bucuresti: ALL; 1993: 142-156.

16. Gordon RE, Gordon KK, Gordon HL, Hursch CJ, Reed KG.  Predicting postnatal emotional adjustment with psychosocial and hormonal measures in early pregnancy. Am J Obstet Gynecol 1986; Jul: 155(1): 80-82.

17. Kittel F, Kornitzer M, De-Backer G, Dramaix M, Sobolski J, Degre S, et all. Type A in relation to job-stress, social and bioclinical variables: the Belgian Physical Fitness Study. J Human Stress 1983; Dec;9(4):37-45.
 

TABLE 1: Relation between basic characteristics and the emergence of obstetrical pathology (n =290)

 TABLE 1 (continue)

TABLE 2: Determinants of pathological pregnancy outcome in multivariate analysis. Presented are odds ratio and their 95% CI.

TABLE 3: Top five life events associated with a high level of stress during pregnancy (stress test)

TABLE 4: Perceived causes of stress during the pregnancy