Maternal age and mortality

A J-shaped curve of the association of MMR and age has been consistently described in the literature, although there are some variations among studies . There is a very clear increase in the risk of maternal death in women older than 30 years. The risk increases with the increasing age of women. However, an increase in risk for young women (<20 years) is also observed, though less consistently.

A community-based survey and case–control study conducted in 25,926 households in 411 villages of the desert districts in Rajasthan, India, evaluated MMR and predictors of maternal deaths. Young age at child birth and poverty were independently associated with increased risk of maternal death . In contrast, a study based on data from the Finnish birth registry did not find any difference in outcomes between women younger than 20 years and those between 20 and 34 years .

Other much larger studies, analysing data from many countries from DHSs or other databases, have reported that while there is a markedly higher risk of maternal death after age 30, the high risk among adolescents is either of a much lower magnitude than is generally assumed or that there was no increased risk of maternal adverse outcomes among adolescents compared with adults .

The risk associated with young age appears to vary depending on the characteristics of the population and age distribution within the under 20 group. While in Finland most women in this group were 18 or 19 years of age, in India, the subgroup under 18 was much larger. In general, however, it appears that the risk associated with younger age is more related to socioeconomic than to physiological factors, except in girls below age 15, a group rarely accounted for in DHS . The association of older maternal age with higher MMR is probably the result of the higher incidence of other coincidental clinical conditions among older women. A review of published studies found that at ≥45 years of age, women have increased rates of pre-existing hypertension and pregnancy complications such as gestational diabetes, gestational hypertension and pre-eclampsia. However, the review also found a trend towards favourable outcomes, even at extremely advanced maternal age (50–65 years), for healthy women who had been screened to exclude pre-existing disease .

Parity and mortality

The association between parity and maternal mortality, although identified in several studies and frequently underestimated by a number of authors, is not as clearly documented as the association with maternal age. A good example that illustrates this unclear association between parity and maternal mortality is given by a historical comparison of the correlation between age and parity with maternal mortality in Sweden from 1950 to 1980 . There was a dramatic reduction in the age and parity-adjusted MMR from 57.2 in the 1951–1960 decade to 7.0 in the 1971–1980 decade. The age-adjusted MMR was the highest among women of parity 1, lowest among women of parity 2–3, and almost as high for women of parity > 4 as for the lowest parity group between 1951 and 1960. Between 1961 and 1970, the age-adjusted MMR remained highest in the lowest parity group and threefold lower in the group with parity 2–3; age-adjusted MMR among women of parity ≥ 4 during this period was almost as low as that of the latter group. Finally, between 1971 and 1980, age-adjusted MMRs among women with parity 1 and 2–3 were similar (9.6 and 8.3), while the lowest age-adjusted MMR was found among women of parity > 4. Thus, while the higher risk of maternal death related to the first birth was confirmed over time, the difference with the parity groups was reduced as the MMRs decreased to very low levels. In contrast, risks with higher parity evident in the first decade had disappeared by the 1970s. Therefore, it is not surprising that the associations between high parity and maternal mortality in low- and middle-income countries are similar to those observed in Sweden until the mid-20th century.

The parity-specific MMR observed in Sweden in the 1970s is similar to the findings of a study conducted in Michigan, USA, in the same period. No increase was found in the age-adjusted MMR with higher parity, but there was a significantly higher age-adjusted MMR among nulliparous women .

A factor that may affect the lower risk associated to parity 4 or higher in the last decade studied in Sweden may be that the distribution of parities in the ≥4 parity group may cluster close to 4 during the last period, while it may have included a larger proportion of women with much higher parity in earlier years, as it is observed today in many less developed countries.

A World Health Organization (WHO)-led study using cross-sectional information collected from 287,035 women giving birth in 373 healthcare institutions in 24 countries in Africa, Asia and Latin America supports this hypothesis when it found ‘a significantly higher risk of maternal mortality among those who had previously had four or more births compared with those who had had none. However, this relationship was attenuated in the adjusted model, where there was a significantly higher risk of maternal mortality only among those who had previously had nine or more births, compared with those who had had none’. Similarly, in a carefully conducted study that analysed data from a demographic surveillance system run over decades in the Matlab area in Bangladesh, a significantly higher risk of maternal death was found only with parity ≥ 7 .

The association of high parity with maternal mortality appears to be biased by a number of other determinants of maternal morbidity and mortality. High parity was found to be associated with low socioeconomic status and education, smoking and alcohol consumption . Significant association was also found between coverage of maternal and child health services and birth order, even when controlling for poverty .

All these factors may explain at least some of the poor obstetric results observed among grand multiparae, such as pre-eclampsia and maternal anaemia, defined as Hb < 10 g/dL . Other complications such as malpresentations, placenta praevia, and postpartum haemorrhage, identified in different studies, may be directly explained by the repeated number of pregnancies and births .

Maternal death is an uncommon occurrence, and only very large samples can give definitive answers. Nevertheless, independent of the possible intermediary factors, maternal mortality does seem to increase with parity, reaching statistical significance above parity 4 in several studies.

Birth spacing and mortality

The association between birth intervals and child mortality has been the subject of a number of studies, but the relationship between birth spacing and maternal mortality has not been so extensively studied, possibly due to the low frequency of maternal deaths. There is a plausible causal mechanism for the association between birth spacing and maternal mortality. A short time interval between the end of a pregnancy and the beginning of the next may cause some form of nutritional depletion, mostly anaemia, with or without previous excessive blood loss. Anaemic women may be at an increased risk of puerperal sepsis, and their tolerance to blood loss may be reduced, which increases the risk of maternal mortality. Probably because the correlation between birth spacing and child mortality is very clear, a similar association with maternal mortality is frequently also underestimated .

Such direct association, however, is not that clear and rather doubtful. A nested case–control study conducted on a cohort of women under demographic surveillance in Matlab, Bangladesh, found that the length of the preceding birth-to-conception interval did not affect the risk of maternal mortality .

A different conclusion was reached by a cross-sectional study based on the Perinatal Information System database of the Latin American Centre for Perinatology, which included 456,889 parous women with singleton infants. This study found that women with inter-pregnancy intervals of ≤5 months had higher risks of maternal death (odds ratio 2.54; 95% CI 1.22–5.38) compared with those conceiving at 18 to 23 months after the previous childbirth . A more recent systematic review, however, concluded that the association between short intervals and maternal death was less clear .

These conflicting results do not confirm the common assumption that very short inter-pregnancy intervals carry a higher risk of maternal mortality and that by increasing such interval, it is possible to reduce MMR . However, even if birth interval does not have a significant impact on mortality, there are several other health reasons that promote longer birth intervals, although not as long as ≥5 years .

Maternal mortality among women with other risk factors

A number of factors influence the risk of maternal death, particularly in low- and middle-income countries, including low education, not having a partner, living in rural areas, distance to the closest health facility, quality of the roads, availability of transportation, attendance for antenatal care, place of delivery, and care provider, if any. . Unfortunately, family planning and contraception cannot change any of these factors.

An important factor that determines maternal death is the presence of pathological conditions that are aggravated by pregnancy and cause (indirect) maternal deaths. A community-based case–control study was conducted in Tigray, Ethiopia, where all maternal deaths between May 2012 and September 2013 were selected as cases and a random sample of mothers who gave birth in the same communities were selected as controls. This study found that women with a history of other illness had the highest risk of maternal death (OR 5.58, 95% CI 2.17–14.30 .

Majority of these women are multiparous and could be identified if they attend a health facility to give birth. The problem is that even women who have the highest risk for death in the next pregnancy are not provided with effective contraception in the postpartum period, even when there is a golden opportunity to offer those services. Most of them leave the hospital without a modern contraceptive method, not only in low- and middle-income countries, but also in some high-income countries.

Maternal age and mortality

A J-shaped curve of the association of MMR and age has been consistently described in the literature, although there are some variations among studies . There is a very clear increase in the risk of maternal death in women older than 30 years. The risk increases with the increasing age of women. However, an increase in risk for young women (<20 years) is also observed, though less consistently.

A community-based survey and case–control study conducted in 25,926 households in 411 villages of the desert districts in Rajasthan, India, evaluated MMR and predictors of maternal deaths. Young age at child birth and poverty were independently associated with increased risk of maternal death . In contrast, a study based on data from the Finnish birth registry did not find any difference in outcomes between women younger than 20 years and those between 20 and 34 years .

Other much larger studies, analysing data from many countries from DHSs or other databases, have reported that while there is a markedly higher risk of maternal death after age 30, the high risk among adolescents is either of a much lower magnitude than is generally assumed or that there was no increased risk of maternal adverse outcomes among adolescents compared with adults .

The risk associated with young age appears to vary depending on the characteristics of the population and age distribution within the under 20 group. While in Finland most women in this group were 18 or 19 years of age, in India, the subgroup under 18 was much larger. In general, however, it appears that the risk associated with younger age is more related to socioeconomic than to physiological factors, except in girls below age 15, a group rarely accounted for in DHS . The association of older maternal age with higher MMR is probably the result of the higher incidence of other coincidental clinical conditions among older women. A review of published studies found that at ≥45 years of age, women have increased rates of pre-existing hypertension and pregnancy complications such as gestational diabetes, gestational hypertension and pre-eclampsia. However, the review also found a trend towards favourable outcomes, even at extremely advanced maternal age (50–65 years), for healthy women who had been screened to exclude pre-existing disease .

Parity and mortality

The association between parity and maternal mortality, although identified in several studies and frequently underestimated by a number of authors, is not as clearly documented as the association with maternal age. A good example that illustrates this unclear association between parity and maternal mortality is given by a historical comparison of the correlation between age and parity with maternal mortality in Sweden from 1950 to 1980 . There was a dramatic reduction in the age and parity-adjusted MMR from 57.2 in the 1951–1960 decade to 7.0 in the 1971–1980 decade. The age-adjusted MMR was the highest among women of parity 1, lowest among women of parity 2–3, and almost as high for women of parity > 4 as for the lowest parity group between 1951 and 1960. Between 1961 and 1970, the age-adjusted MMR remained highest in the lowest parity group and threefold lower in the group with parity 2–3; age-adjusted MMR among women of parity ≥ 4 during this period was almost as low as that of the latter group. Finally, between 1971 and 1980, age-adjusted MMRs among women with parity 1 and 2–3 were similar (9.6 and 8.3), while the lowest age-adjusted MMR was found among women of parity > 4. Thus, while the higher risk of maternal death related to the first birth was confirmed over time, the difference with the parity groups was reduced as the MMRs decreased to very low levels. In contrast, risks with higher parity evident in the first decade had disappeared by the 1970s. Therefore, it is not surprising that the associations between high parity and maternal mortality in low- and middle-income countries are similar to those observed in Sweden until the mid-20th century.

The parity-specific MMR observed in Sweden in the 1970s is similar to the findings of a study conducted in Michigan, USA, in the same period. No increase was found in the age-adjusted MMR with higher parity, but there was a significantly higher age-adjusted MMR among nulliparous women .

A factor that may affect the lower risk associated to parity 4 or higher in the last decade studied in Sweden may be that the distribution of parities in the ≥4 parity group may cluster close to 4 during the last period, while it may have included a larger proportion of women with much higher parity in earlier years, as it is observed today in many less developed countries.

A World Health Organization (WHO)-led study using cross-sectional information collected from 287,035 women giving birth in 373 healthcare institutions in 24 countries in Africa, Asia and Latin America supports this hypothesis when it found ‘a significantly higher risk of maternal mortality among those who had previously had four or more births compared with those who had had none. However, this relationship was attenuated in the adjusted model, where there was a significantly higher risk of maternal mortality only among those who had previously had nine or more births, compared with those who had had none’. Similarly, in a carefully conducted study that analysed data from a demographic surveillance system run over decades in the Matlab area in Bangladesh, a significantly higher risk of maternal death was found only with parity ≥ 7 .

The association of high parity with maternal mortality appears to be biased by a number of other determinants of maternal morbidity and mortality. High parity was found to be associated with low socioeconomic status and education, smoking and alcohol consumption . Significant association was also found between coverage of maternal and child health services and birth order, even when controlling for poverty .

All these factors may explain at least some of the poor obstetric results observed among grand multiparae, such as pre-eclampsia and maternal anaemia, defined as Hb < 10 g/dL . Other complications such as malpresentations, placenta praevia, and postpartum haemorrhage, identified in different studies, may be directly explained by the repeated number of pregnancies and births .

Maternal death is an uncommon occurrence, and only very large samples can give definitive answers. Nevertheless, independent of the possible intermediary factors, maternal mortality does seem to increase with parity, reaching statistical significance above parity 4 in several studies.

Birth spacing and mortality

The association between birth intervals and child mortality has been the subject of a number of studies, but the relationship between birth spacing and maternal mortality has not been so extensively studied, possibly due to the low frequency of maternal deaths. There is a plausible causal mechanism for the association between birth spacing and maternal mortality. A short time interval between the end of a pregnancy and the beginning of the next may cause some form of nutritional depletion, mostly anaemia, with or without previous excessive blood loss. Anaemic women may be at an increased risk of puerperal sepsis, and their tolerance to blood loss may be reduced, which increases the risk of maternal mortality. Probably because the correlation between birth spacing and child mortality is very clear, a similar association with maternal mortality is frequently also underestimated .

Such direct association, however, is not that clear and rather doubtful. A nested case–control study conducted on a cohort of women under demographic surveillance in Matlab, Bangladesh, found that the length of the preceding birth-to-conception interval did not affect the risk of maternal mortality .

A different conclusion was reached by a cross-sectional study based on the Perinatal Information System database of the Latin American Centre for Perinatology, which included 456,889 parous women with singleton infants. This study found that women with inter-pregnancy intervals of ≤5 months had higher risks of maternal death (odds ratio 2.54; 95% CI 1.22–5.38) compared with those conceiving at 18 to 23 months after the previous childbirth . A more recent systematic review, however, concluded that the association between short intervals and maternal death was less clear .

These conflicting results do not confirm the common assumption that very short inter-pregnancy intervals carry a higher risk of maternal mortality and that by increasing such interval, it is possible to reduce MMR . However, even if birth interval does not have a significant impact on mortality, there are several other health reasons that promote longer birth intervals, although not as long as ≥5 years .

Maternal mortality among women with other risk factors

A number of factors influence the risk of maternal death, particularly in low- and middle-income countries, including low education, not having a partner, living in rural areas, distance to the closest health facility, quality of the roads, availability of transportation, attendance for antenatal care, place of delivery, and care provider, if any. . Unfortunately, family planning and contraception cannot change any of these factors.

An important factor that determines maternal death is the presence of pathological conditions that are aggravated by pregnancy and cause (indirect) maternal deaths. A community-based case–control study was conducted in Tigray, Ethiopia, where all maternal deaths between May 2012 and September 2013 were selected as cases and a random sample of mothers who gave birth in the same communities were selected as controls. This study found that women with a history of other illness had the highest risk of maternal death (OR 5.58, 95% CI 2.17–14.30 .

Majority of these women are multiparous and could be identified if they attend a health facility to give birth. The problem is that even women who have the highest risk for death in the next pregnancy are not provided with effective contraception in the postpartum period, even when there is a golden opportunity to offer those services. Most of them leave the hospital without a modern contraceptive method, not only in low- and middle-income countries, but also in some high-income countries.