SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Objective  Some data suggest an association between teenage childbearing and premature death. Whether this possible increase in risk is associated with social circumstances before or after childbirth is not known. We studied premature death in relation to age at first birth, social background and social situation after first birth.

Design  Population-based cohort study.

Setting  Women born in Sweden registered in the 1985 Swedish Population Census.

Population  Swedish women born 1950–1964 who had their first infant before the age of 30 years (N= 460,434).

Methods  Information on the women's social background and social situation after first birth was obtained from Population Censuses. The women were followed up with regard to cause of death from December 1, 1990 to December 31, 1995. Mortality rate ratios and 95% confidence intervals (CI) were calculated.

Main outcome measures  Mortality rates by cause of death.

Results  Independent of socio-economic background, teenage mothers faced an increased risk of premature death later in life compared with older mothers (rate ratio 1.6, 95% CI 1.4–1.9). The increased risk was most evident for deaths from cervical cancer, lung cancer, ischaemic heart disease, suicide, inflicted violence and alcohol-related diseases. Some, but not all, of these increases in risk were associated with the poorer social position of teenagers mothers.

Conclusions  Teenage mothers, independent of socio-economic background, face an increased risk of premature death. Strategies to reduce teenage childbearing are likely to contribute to improved maternal and infant health.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Compared with older mothers, teenage mothers are reported to have poorer self-reported health status1 and higher overall mortality later in life.2 There is also evidence of increased risks of specific diseases among women who have their first birth at young age, particularly heart disease3,4 and cervical cancer.5,6 It has been assumed that these risks may be due to the social background rather than biological factors.4 Teenage mothers have a less favourable social background than older mothers,7–10 and social background has not been allowed in previous studies.1–6 Teenage motherhood may also, independent of family background, increase the risk of a being in a less advantageous social situation later in life, including increased risks of low socio-economic status, unemployment, lone motherhood and welfare dependency.11 Socio-economic situation, marital status and lone motherhood have been associated with morbidity and mortality among women.12–16 Therefore, a possible association between low maternal age at first birth and subsequent risk of premature death may be the effect of the social circumstances created by early childbirth.

The aim of the present study is to investigate the relationship between maternal age at first birth and risks of total, as well as cause-specific premature mortality (26–45 years). We will also investigate to what extent possible differences in mortality are independent of women's social background and are mediated by socio-economic factors after childbirth.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Our study is a follow up of all women born in Sweden from 1950 to 1964, registered in the 1985 Swedish Population and Housing Census, who had their first infant between 1964 and 1989 before the age of 30 years. We excluded women who emigrated (n= 196) or died (n= 796) between the time of the 1985 Census and the time for the start of the follow up (December 1, 1990). The study population comprised 460,343 Swedish women, alive and living in Sweden on December 1, 1990.

Information on father and mother's socio-economic status in 1960 was obtained by record linkage to the 1960 Swedish Census. We identified the parents through individual record linkage between the study population and the Population Register. The record linkages were possible by use of the unique national registration numbers assigned to each Swedish resident. In all, we received information on socio-economic background for 97% of the study population. We used information on both the fathers' and the mothers' socio-economic position and formed one variable based on the person in the household that, at the time of the Census, ranked the highest.

Information about socio-economic situation and family situation later in life was obtained from the 1990 Population Census, when the women were between 26 and 40 years. The classification of socio-economic situation of the women and their parents is based on occupation. Different occupations are allocated to different socio-economic groups based primarily on the degree of education required for each specific occupation.17 By individual record linkage to the Population Register, covering all births in Sweden, we were able to assess the number of infants for each woman from 1964 through 1993. Information on welfare dependency in 1990 on all nationally registered persons was obtained from the Total Enumeration Income Survey held by Statistics Sweden, a nationwide register with information on social benefits.

The following factors were treated categorically in the analyses: age at first birth, socio-economic background according to household, own socio-economic status, family situation, number of infants and welfare dependency. Age at the start of the follow up (December 1, 1990) was treated as a continuous variable in all multivariate analyses. Age at first delivery was stratified into 17 years or less, 18–19, 20–24 and 25–29 years. Background socio-economic situation was categorised into blue-collar worker, white-collar worker, self-employed (including farmers) and not gainfully employed. Family situation was dichotomised into living as a single person or not, irrespective of marital status. Parity was arranged into three groups: one, two to three, and four infants or more. Information on welfare dependency was dichotomised as to whether any person in the household received social allowance or not any time during the year 1990.

We linked all women to the Causes of Death Registry, held by the National Board of Health and Welfare. We identified deaths from December 1, 1990 to December 31, 1995 with information on date and cause of death. All deaths were coded according to the International Classification of Diseases, ninth revision (ICD-9). Cause of death was first analysed using the main chapters of underlying cause of death. Main causes of death that were significantly associated with maternal age at first birth were then further subdivided into specific underlying causes. Neoplasms (ICD-9: 140–239) were subdivided into malignant neoplasms of the trachea, bronchus and lung (ICD-9: 162), malignant neoplasm of breast (ICD-9: 174), malignant neoplasm of cervix uteri (ICD-9: 180), and other neoplasms. Diseases of the circulatory system (ICD-9: 390–459) were subdivided into ischaemic heart disease (ICD-9: 410–414), cerebrovascular diseases (ICD-9: 430–438) and other circulatory diseases. External causes of death (ICD-9: E800–E999) were subdivided into suicide (ICD-9: E950–E959), violence, assault and homicide (ICD-9: E960–E969), motor vehicle accidents (ICD-9: E810–E819), accidental poisoning by drugs, medicaments and biologicals (ICD-9: E850–E858) and other external causes. Alcohol-related mortality was defined as cases with ICD-9 codes 291, 303, 305.0, 357.5, 425.5, 535.3, 571.0–571.3, E860, E980 in combination with 980 as underlying or contributing cause of death.18

We counted the person-years at risk for individuals that were alive and living in the country during the period of follow up (December 1, 1990 to December 31, 1995). The month of death was counted as 15 days, and the year of emigration was counted as six months. Poisson regression analyses were performed to estimate the effect of age at first birth on mortality rate ratios, using women aged 20–29 years at first birth as the reference group. The mortality rate ratios were primarily adjusted for age in 1990 and background socio-economic situation. We also in a final step adjusted for socio-economic characteristics after childbirth, including socio-economic status, family situation, number of infants and welfare dependency.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Among women having their first infant before aged 30 years, 13% were teenage childbearers. Compared with older mothers, teenage mothers more often belonged to older birth cohorts, and were more often from households with blue-collar worker or unemployed parents (Table 1). Later in life, teenage mothers were more often blue-collar workers, not gainfully employed, more often lived without a partner, were dependent on social welfare and also more often had four infants or more (Table 1). The large differences in parity distribution with regard to age at first birth were evident in all maternal birth cohorts (data not shown).

Table 1.  Distribution of characteristics (in %) according to maternal age at first birth among women born 1950–1964 in Sweden.
 Maternal age at first birth (years)
≤17 (n= 14,158)18–19 (n= 46,528)20–24 (n= 218,479)25–29 (n= 181,269)
  • *

    In 1990.

  • Births from 1964 to 1993.

Birth cohort (age in 1990)
1950–1954 (36–40)49.345.339.535.2
1955–1959 (31–35)34.734.731.637.5
1960–1964 (26–30)16.120.028.927.3
 
Socio-economic background
Blue-collar worker60.757.849.939.6
White-collar worker18.221.429.538.9
Self-employed/farmer12.513.314.816.8
Not gainfully employed4.33.82.92.2
Unclassifiable/missing data4.43.83.02.5
 
Socio-economic status*
Blue-collar worker53.454.748.331.2
White-collar worker21.722.332.752.0
Self-employed/farmer3.33.03.32.8
Not gainfully employed14.513.310.48.9
Unclassifiable/missing data7.26.75.45.1
 
Family situation*
Single25.222.917.011.6
Not single69.973.080.285.8
Missing data4.94.02.82.6
 
No. of infants
113.712.812.717.2
2–364.769.978.679.7
4+21.517.38.73.2
 
Welfare dependency*
Yes16.313.77.73.8
No83.786.392.396.2

In all, 1269 of the 460,343 women died during the period of follow up, and the age at death ranged from 30 to 45 years. The risk of premature death decreased with increasing maternal age at first birth, from 107 per 100,000 person-years at risk among mothers aged 17 years or less, to 41.8 among mothers aged 25–29 years (Table 2). Compared with mothers aged 20–24 years at first birth, mothers aged 17 years or less at first birth had a 70% increase in risk of premature death and those aged 18–19 years, a 50% risk increase. Risk of premature death was also substantially increased for women whose parents were not gainfully employed, for unemployed women, single women, women with only one child, and a more than fourfold increase in risk for premature death was observed among women receiving social benefits. Women who were unclassifiable or had missing data on socio-economic position and family situation also had substantially increased death rates.

Table 2.  Characteristics among women born 1950 to 1964 in Sweden in relation to mortality.*
 No. of deathsMortality rate no./100,000 person-yearsMortality rate ratio95% CI
  • *

    Deaths from December 1, 1990 to December 31, 1995.

  • Adjusted for age in 1990.

  • Reference group.

  • §

    In 1990.

  • #

    Births from 1964 to 1993.

Maternal age at first birth
≤1777107.01.71.4–2.2
18–1920586.61.51.3–1.7
20–2460154.01.0
25–2938641.80.80.7–0.9
 
Socio-economic background
Blue-collar worker61656.01.21.0–1.4
White-collar worker34245.51.0
Self-employed/farmer20757.51.10.9–1.3
Not gainfully employed5280.41.91.4–2.6
Unclassifiable/missing data5276.01.30.99–1.8
 
Socio-economic status§
Blue-collar worker46046.31.21.0–1.3
White-collar worker42446.51.0
Self-employed/farmer2332.20.70.4–1.0
Not gainfully employed26639.72.92.5–3.4
Unclassifiable/missing data9672.22.01.6–2.5
 
Family situation§
Single35195.32.01.9–2.4
Not single82443.21.0
Missing data94138.13.52.8–4.3
 
Number of infants#
131799.32.01.8–2.3
2–385546.91.0
4+9753.31.00.8–2.3
 
Welfare dependency§
Yes274166.14.33.8–4.9
No99545.71.0
 
Total mortality126954.1  

Among women who had their first infant during the teenage period, the mortality was 91.4 per 100,000 person-years, compared with 48.5 per 100,000 among older mothers. The most common cause of death among both teenage mothers and older mothers was neoplasm, followed by injury and poisoning, and diseases of the circulatory system (Table 3).

Table 3.  Number of deaths and mortality rates by cause of death in relation to maternal age at first birth among women born in Sweden 1950–1964.*
Cause of deathMaternal age at first birth (years)
11–1920–29
nn/100,000 person-yearsnn/100,000 person-years
  • *

    Deaths from December 1, 1990 to December 31, 1995.

  • Including diseases of the blood and blood forming organs (ICD-9: 280–289), diseases of the genitourinary system (ICD-9: 580–626), complications of pregnancy, childbirth and the puerperium (ICD-9: 630–676), diseases of the musculoskeletal system and connective tissue (ICD-9: 710–739), congenital anomalies (ICD-9: 740–759), and symptoms, signs and ill-defined conditions (ICD-9: 780–789).

Infectious and parasitic diseases, ICD-9: 001–13931.0120.6
Neoplasms, ICD-9: 140–23910233.144321.8
Endocrine, nutrition and metabolic diseases, and immunity disorders, ICD-9: 240–27941.3221.1
Mental disorders, ICD-9: 290–31972.3190.9
Diseases of the nervous system and sense organs, ICD-9: 320–38992.9291.4
Diseases of the circulatory system, ICD-9: 390–4594113.31105.4
Diseases of the respiratory system, ICD-9: 460–51992.9190.9
Diseases of the digestive system, ICD-9: 520–579113.6271.3
Injury and poisoning, ICD-9: E800–E9999430.527913.7
Other causes20.6271.3
Total mortality28291.498748.5

Compared with women aged 20–29 years at first birth, teenage mothers faced a 30% increase in risk of dying from cancer, while risks of death due to diseases in the circulatory system, the respiratory and digestive systems, injury and poisoning were all more than doubled (Table 4). Adjustment for socio-economic background did not essentially change the age-related mortality risks for neoplasms, circulatory diseases, injury and poisoning, while mortality risks of diseases of the respiratory and digestive systems decreased and were no longer significant (Table 4).

Table 4.  Mortality rate ratios with 95% CI by cause of death in relation to young maternal age at first birth (≤19 years) among women born in Sweden 1950 to 1964.* Reference group: maternal age 20-29 at first birth.
Cause of deathMortality rate ratio95% CIMortality rate ratio95% CIMortality rate ratio§95% CI
  • *

    Deaths from December 1, 1990 to December 31, 1995.

  • Adjusted for age in 1990.

  • Adjusted for age in 1990 and background socio-economic status in 1960.

  • §

    Adjusted for age in 1990, background socio-economic status, and socio-economic position, family situation, welfare dependency in 1990 and number of infants.

Neoplasms1.31.1–1.71.31.1–1.61.31.1–1.7
Diseases of the circulatory system2.21.5–3.12.11.5–3.01.81.2–2.6
Diseases of the respiratory system2.81.3–6.31.60.7–3.61.20.5–2.7
Diseases of the digestive system2.31.1–4.61.50.7–3.11.20.6–2.5
Injury and poisoning2.11.6–2.62.01.6–2.61.61.2–2.0
Total mortality1.71.5–1.91.61.4–1.91.41.2–1.6

In order to investigate whether the age-related mortality risks could be explained by events after childbirth, variables reflecting social situation later in life, such as socio-economic situation, family situation, dependence on social welfare and number of infants, were included in the analyses. These adjustments further decreased the excess risks among teenage mothers for overall mortality, cardiovascular mortality and mortality caused by injury or poisoning, while risk of dying from cancer remained at the same level (Table 4).

The major causes of death associated with teenage childbearing (neoplasms, circulatory diseases, injury and poisoning) were divided into more specific causes of death (Table 5). After adjustment for pre-childbearing socio-economic status, teenage mothers faced, compared with women aged 20–29 years at first birth, a more than doubled risk of dying from cervical cancer, and a doubled risk of dying from lung cancer. Teenage mothers also faced an almost 10-fold increase in risk for death due to inflicted violence, while risks for death due to ischaemic heart disease, diseases of the circulatory system other than cerebrovascular diseases, suicide and alcohol-related mortality were doubled or more than doubled. When the social situation after first childbirth was also taken into account, the mortality rate ratios for most causes of death decreased. However, the increase in risk of premature death among teenage childbearers due to cervical cancer, ischaemic heart disease, suicide and inflicted violence remained significantly increased (Table 5).

Table 5.  Mortality rate ratios with 95% CI by cause of death in relation to young maternal age at first birth (≤19 years) among women born in Sweden 1950 to 1964.* Reference group: maternal age 20–29 years at first birth.
Cause of deathMortality rate ratio95% CIMortality rate ratio95% CI
  • *

    Deaths from December 1, 1990 to December 31, 1995.

  • Adjusted for age in 1990.

  • Adjusted for age in 1990, background socio-economic status in 1960, and socio-economic position, family situation, welfare dependency in 1990 and number of infants.

Neoplasms
Breast0.90.6–1.41.00.6–1.5
Cervix2.61.4–5.02.41.2–4.6
Lung2.01.0–3.91.80.9–3.5
Other1.30.96–1.81.30.98–1.8
 
Diseases of the circulatory system
Ischaemic heart disease2.81.5–5.42.21.2–4.3
Cerebrovascular disease1.50.8–2.91.40.7–2.7
Other2.21.2–4.11.80.9–3.4
 
Injury and poisoning
Suicide1.91.4–2.61.51.1–2.0
Violence, assault, homicide9.83.2–29.85.71.8–17.9
Motor vehicle accidents1.50.8–2.91.50.7–2.9
Accidental poisoning2.50.6–9.41.80.5–7.2
Other1.80.9–3.81.50.7–3.3
 
Alcohol-related mortality2.21.5–3.21.40.9–2.0

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

We found convincing evidence that teenage childbearers face an increased risk of premature death independent of socio-economic background. The age-related increase in risk was most evident for deaths related to a health damaging lifestyle, poor psychosocial health or a violent environment. The poorer social situation after childbirth among teenage childbearers explained some but not all of these risks: risks remained significantly increased for overall mortality, mortality due to cervical cancer, ischaemic heart disease, suicide and inflicted violence.

The strength of the present population-based investigation is the large sample size, giving consistent results with high precision. More than 460,000 women and more than 1200 deaths in the age span from 30 to 45 years were included, which makes it possible to also study cause-specific mortality. Furthermore, the population-based study design ensures that no selection bias is present, and information on pre-childbearing characteristics and social situation later in life was recorded independent of the outcome studied.

Socio-economic situation before first childbirth was assessed through information on parents' socio-economic situation in 1960. As the included women were born from 1950 to 1964, the oldest were 10 years and some were not yet born when background socio-economic situation was measured. The women were also at different ages when we assessed their social situation in adulthood (26–40 years). Adjustment for age in the analyses should minimise, but cannot exclude, some misclassification of socio-economic situation before and after first birth. Number of infants was to some extent collected during the time of follow up, since this started in December 1990 and we counted births until the end of 1993. There were, however, relatively few deaths and the parity distribution among women in the birth cohorts included were essentially the same in 1990 and in 1993.

In accordance with previous investigations,19 we found that teenage mothers tend to come from less advantaged environments. Although mothers with a blue-collar background and those having unemployed parents were at increased risk of premature death, low age at first birth increased the risk for premature death independent of background socio-economic situation. Independent of social background, teenage childbearers also face a less advantageous socio-economic situation later in life, are more often single mothers and have more children than older childbearers.11 However, even after taking these circumstances into account, teenage childbearers face a 40% greater risk for premature death than older mothers. This excess risk is probably due to lifestyle factors and being in a vulnerable life situation after childbirth, which seems to be independent of the women's social situation.

The increase in risk of premature death among teenage childbearers was most pronounced for deaths associated with lifestyle factors, psychosocial health or violent environment. We find it unlikely that these increased risks are effects of age at first birth itself, but rather caused by unmeasured lifestyle related factors. Teenage mothers in Sweden have a much higher proportion of daily smokers during pregnancy than older mothers.19 Smoking may also indicate other less advantageous health habits,20 and teenage childbearers in this study were also at increased risk of alcohol-related mortality. Smoking increases the risk for not only lung cancer, but also ischaemic heart disease,21,22 and probably also cervical cancer.22 It has been suggested that smoking is a major risk factor for cardiovascular disease primarily among young women,23 and the results from the present investigation are consistent with findings from a case–control study showing an increase in risk for myocardial infarction4 and coronary heart disease3 among women with a first birth at an early age. Other important risk factors for cardiovascular disease among women is hypertension, hypercholesterolemia, overweight and stress.21,22 However, we lack information on how these risk factors are distributed according to age at first birth among Swedish women.

Early age at first intercourse, an early first birth and a high number of sexual partners are associated with increased risks of human papillomavirus,24,25 the main cause of cervical cancer. The risk of dying from cervical cancer was more than doubled. In a previous case–control study, being 15 years or less at first birth was associated with a fourfold increase in risk for cervical cancer.5 A population-based case–control study also showed a clear effect of low age at fist birth (<21 years) on risk for cervical cancer after controlling for socio-demographic factors.6

Compared with older mothers, teenage mothers face a 10-fold increase in risk of death caused by inflicted violence, indicating that young mothers probably also run a substantially increased risk of being subjected to violence with non-fatal outcome. Teenage mothers have previously been reported to be at increased risk for homicide postpartum.26 Risk factors for women at home of death due to inflicted violence of women in the home are: suggested to be living alone, illicit drug use by any member of the household and prior domestic violence.27 Although there were few deaths caused by inflicted violence during the five years under study (seven teenage childbearers and five older mothers), the results indicate that teenage childbearers are exposed to a more violent environment—violent men—than older childbearers.

There is an association between socio-economic situation and risk of suicide.28,29 In the present investigation, the association between teenage childbearing and suicide was to some extent an effect of a less advantaged social situation. However, also after taking socio-economic circumstances later in life into account, teenage mothers faced a 50% higher risk for suicide than older mothers. The present study lacks information about depressive disorders, adjustment disorders and substance abuse, all of which are important risk factors for suicide.30

Teenage childbearers, independent of socio-economic background, face an increased risk of premature death, most evident for deaths from cervical cancer, lung cancer, ischaemic heart disease, suicide, inflicted violence and alcohol-related diseases. Experiencing a teenage pregnancy will certainly influence subsequent risks of low socio-economic status, unemployment and being a single mother. The less favourable social situation of teenage mothers may in turn predispose to a social vulnerability and unhealthy lifestyle. Strategies to reduce teenage childbearing are likely to contribute to improved maternal and infant health.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

The study was supported by grants from The Swedish Council for Social Research (Project FO265/1998).

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  • 1
    Kington R, Lillard L, Rogowski J. Reproductive history, socio-economic status, and self-reported health status of women aged 50 years or older. Am J Public Health 1997;87: 3337.
  • 2
    Lund E, Arnesen E, Borgan J-K. Pattern of childbearing and mortality in married women—a national prospective study from Norway. J Epidemiol Community Health 1990;44: 237240.
  • 3
    Beard CM, Fuster V, Annegers JF. Reproductive history in women with coronary heart disease. A case–control study. Am J Epidemiol 1984;120: 108114.
  • 4
    Palmer JR, Rosenberg L, Shapiro S. Reproductive factors and risk of myocardial infarction. Am J Epidemiol 1992;136: 408416.
  • 5
    Bosch FX, Munoz N, de Sanjose S, et al. Risk factors for cervical cancer in Columbia and Spain. Int J Cancer 1992;52: 750758.
  • 6
    Bjorge T, Kravdal O. Reproductive variables and risk of uterine cervical cancer in Norwegian registry data. Cancer Causes Control 1996;7: 351357.
  • 7
    Hoffman SD, Foster EM, Furstenberg Jr FF. Reevaluating the costs of teenage childbearing. Demography 1993;30: 113.
  • 8
    Geronimus AT, Korenman S. The socio-economic costs of teenage childbearing: evidence and interpretation. Demography 1993;30: 281290.
  • 9
    Steven-Simons C, Lowy R. Teenage childbearing. An adaptive strategy for the socioeconomically disadvantaged or a strategy for adapting to socio-economic disadvantage? Arch Pediatr Adolesc Med 1995;149: 912915.
  • 10
    Spivak H, Weitzman M. Social barriers faced by adolescent parents and their children. JAMA 1987;258: 15001504.
  • 11
    Otterblad Olausson P, Haglund B, Ringbäck Weitoft G, Cnattingius S. Teenage childbearing and long-term socio-economic consequences: a case study in Sweden. Fam Plann Perspect 2001;33: 7074.
  • 12
    Mackenbach JP, Kunst AE, Calvelaars AEJM, Groenhof F, Geurts JMJ. EU working group on socio-economic inequalities in health. Socio-economic inequalities in morbidity and mortality in Western Europe. Lancet 1997;349: 16551659.
  • 13
    Sundqvist J, Johansson SE. Indicators of socio-economic position and their relation to mortality in Sweden. Soc Sci Med 1997;45: 17571766.
  • 14
    Benzeval M. The self-reported health status of lone parents. Soc Sci Med 1998;46: 13371353.
  • 15
    Popey J, Jones G. Patterns of health and illness amongst lone parents. J Soc Pol 1990;19: 499534.
  • 16
    Ringbäck Weitoft G, Haglund B, Rosén M. Mortality among lone mothers in Sweden: a population study. Lancet 2000;355: 12151219.
  • 17
    Statistics Sweden. Swedish socio-economic classification. Reports on Statistical Co-ordination 1982, 4. Örebro: Statistics Sweden, 1984 [in swedish with an English summary].
  • 18
    The National Board of Health and Welfare, Centre for Epidemiology. Causes of Death 1996. Stockholm: The National Board of Health and Welfare, 1998.
  • 19
    Otterblad Olausson PM, Cnattingius S, Goldenberg RL. Determinants of poor pregnancy outcomes among teenagers in Sweden. Obstet Gynecol 1997;89: 451457.
  • 20
    The National Board of Health and Welfare, Centre for Epidemiology. Sweden's Public Health Report 1997. Stockholm: The National Board of Health and Welfare, 1998.
  • 21
    Wenger NK. Hypertension and other cardiovascular risk factors in women. Am J Hypertens 1995;8: 9499.
  • 22
    Sherman CB. Health effects of cigarette smoking. Clin Chest Med 1991;12: 643658.
  • 23
    Burkman RT. Obesity, stress, and smoking: their role as cardiovascular risk factors in women. Am J Obstet Gynecol 1988;158: 15921597.
  • 24
    TomatisL, AitioA, DayNE et al., editors. Cancer: Causes Occurrence and Control. Lyon: IARCH Scientific Publication No. 100, 1990.
  • 25
    EstèveJ, KrideerA, FerleyJ, ParkinDM, editors. Facts and Figures of Cancer in the European Community. Lyon: IARC Commission of the European Communities, 1993.
  • 26
    Dietz PM, Rochat RW, Thompson BL, Berg CJ, Griffin GW. Differences in risk of homicide and other fatal injuries between postpartum women and women of childbearing age: implications and prevention. Am J Public Health 1998;88: 641643.
  • 27
    Bailey JE, Kellerman AL, Somes GW, Banton JG, Rivara FP, Rushforth NP. Risk factors for violent death of women in the home. Arch Intern Med 1997;157: 777782.
  • 28
    Ferrada-Noli M. Social psychological vs socio-economic hypotheses on the epidemiology of suicide: an empirical study. Psychol Rep 1996;79: 707710.
  • 29
    Johansson LM, Sundqvist J, Johansson SE, Qvist J, Bergman B. The influence of ethnicity and social and demographic factors on Swedish suicide rates. A four year follow-up study. Soc Psychiatry Epidemiol 1997;32: 165170.
  • 30
    Åsgård U. A psychiatric study of suicide among urban Swedish women. Acta Psychiatr Scand 1990;82: 115124.

Accepted 2 April 2004