Abstract
Objective
To examine transition times between long-term conditions (LTCs) and their associated determinants among childbearing women.
Design
A population-based cohort study. We estimated median times to progress from 1 to 2, 2 to 3, 3 to 4 and 4 to ≥5 LTCs, stratified by ethnicity, socioeconomic status and region. 10-year risks of progression from first to second LTC were estimated by entry cohort (pre-2013 vs ≥2013) using Kaplan–Meier methods, with determinants assessed using Weibull survival models.
Setting
English primary care records (CPRD Aurum and GOLD) and pregnancy registers.
Participants
A total of 2 160 157 women with a first LTC diagnosis during reproductive years (ages 15–49) and at least one recorded pregnancy between 2003 and 2022.
Results
Approximately 65% with one LTC progressed to develop two or more LTCs. The median time to develop additional LTCs decreased progressively: 5.58 years (from 1 to 2), 4.45 years (from 2 to 3), 3.89 years (from 3 to 4) and 3.68 years (from 4 to 5+). Women diagnosed with their first LTC from 2013 onwards transitioned to second LTC faster than those diagnosed before 2013 (χ² = 4260.55, p<0.001). Faster progression was observed among women from the most deprived backgrounds compared with the least deprived (TR 0.79, 95% CI 0.79 to 0.80). Compared with Southwest England, women in the Northwest, West Midlands and Southeast acquired additional LTCs more rapidly. Ethnic variations in the risk of progression were also observed in specific contexts. Furthermore, analysis of baseline condition-specific trajectories showed that women with an initial diagnosis of cardiomyopathy experienced the fastest transition to a subsequent LTC (median survival time (MST) 3.89 years; 95% CI 0.99 to 11.78). This was followed by women with coronary heart disease (MST 4.21 years; 95% CI 1.30 to 11.34), diabetes (MST 4.25 years; 95% CI 1.39 to 9.20) and anxiety (MST 4.29 years; 95% CI 1.46 to 10.12).
Conclusion
LTC accumulation among childbearing women is driven by socioeconomic deprivation, ethnicity and regional disparities. Targeted interventions for high-risk groups, alongside efforts to address structural inequalities, may help slow this progression.
Keywords: Multiple Chronic Conditions, Women's Health, Women's Health Services, Reproductive Health, Reproductive Health Services
WHAT IS ALREADY KNOWN ON THIS TOPIC
There is an increasing burden of multiple long-term conditions (LTCs) among younger populations, with more women in particular entering pregnancy with pre-existing conditions.
Multiple LTCs are linked to negative pregnancy outcomes.
Little is known about the pace and drivers of disease accumulation, although understanding these could be key to identifying high-risk groups for targeted monitoring and interventions.
WHAT THIS STUDY ADDS
The time to accumulate additional LTCs is progressively shortening, with more recent cohorts of childbearing women developing multimorbidity more rapidly.
Socioeconomic status, region of residence and ethnicity influence the pace of LTC accumulation.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The observed acceleration in the burden of multiple LTCs, especially in the contemporary women population, may reflect increasing physiological vulnerability, changes in lifestyle, improvements in medical diagnosis or growing challenges in disease management.
Concerted actions targeting modifiable factors and structural health inequalities may help prevent or delay LTC accumulation, especially among high-risk groups.
Introduction
Evidence suggests a rising proportion of women become pregnant with multiple long-term conditions (MLTCs), and a dose–response association between the number of chronic conditions and the risk of adverse pregnancy outcomes.1 2 In general, women with MLTCs are at higher risk of severe maternal complications and adverse outcomes compared with those without such conditions.3 4 Preventing MLTCs in women of reproductive age is critical, given its adverse effects on maternal and infant health, including increased mortality risk, more complex care needs, greater healthcare burden, intergenerational consequences and widened health disparities.1 2 5
However, little is known about the timing of transitions from one LTC to the next among childbearing women and on the factors associated with this progression. This understanding can help clinicians more rapidly identify women at risk of developing MLTCs, support better prepregnancy risk stratification and targeted optimisation of health before conception, inform tailored prevention strategies for modifiable factors (such as deprivation), guide resource allocation and health-service planning and shape policies aimed at reducing health inequalities. Hence, by analysing a cohort of childbearing women in England, we examine the duration of transitions from 1 to 2, 2 to 3, 3 to 4 and 4 to 5 or more LTCs during their reproductive years.
Methods
Study design and population
This is a population-based cohort study. The study population comprised pseudonymised women with at least one LTC first diagnosed during reproductive years (ages 15–49),6 and a recorded pregnancy between 1 January 2003 and 31 December 2022, regardless of the outcome. The cohort was identified from the Clinical Practice Research Datalink (CPRD) Pregnancy Register, linked with primary care records. Eligible women entered the study when they were diagnosed with their first LTC during childbearing age and were followed up until the general practice’s last data collection date, the point at which they reached age 49, or their date of death, whichever occurred first. Since our aim was to exclusively examine the time taken to acquire additional LTCs during the reproductive years starting from a single LTC, we excluded women who: (1) had a prior history of an LTC diagnosed before age 15, (2) were first diagnosed with an LTC after age 49 or (3) had more than one condition (ie, MLTCs) diagnosed simultaneously at onset.
Data source
We linked the Pregnancy Register with the English CPRD AURUM and GOLD datasets. The CPRD databases provide longitudinal electronic health records from UK general practices. CPRD AURUM represents roughly a quarter of the UK population, while CPRD GOLD includes data on about 5%.7 8 The datasets offer detailed information on patient demographics, clinical symptoms, diagnostic tests, medical diagnoses, prescribed treatments and referrals. The data were deduplicated, and the completeness of ethnicity information was improved through linkage with the Hospital Episode Statistics patient table. We also linked the CPRD data with patient-level index of multiple deprivation (IMD) data9—an index derived from various indicators of material deprivation across multiple domains, including income, employment, education and skills, health, housing, crime, access to services and the living environment—to assess socioeconomic status. In total, data from 2 160 157 women were included in the final analysis (online supplemental figure S1).
Procedure
Outcome of interest
The primary outcome was the time to accrual of additional LTCs from 1 to 2, 2 to 3, 3 to 4 and 4 to 5+LTCs during the childbearing years. LTCs were defined based on a list of 79 chronic conditions developed by the MuM-PreDiCT Consortium through literature review and public and patient engagement (online supplemental table S1).10 The list was established considering condition prevalence, potential impact on pregnancy and the perceived importance of the conditions by women.10 We adapted this list because of its relevance to our study population. These conditions were identified in primary care records using READ and Systematized Nomenclature of Medicine—Clinical Terms (SNOMED CT) codes.11
Main exposures
The main exposure was the first LTC diagnosed during the childbearing years, which served as the baseline for examining transition times between LTCs. For later transitions, each new exposure was defined as the LTC diagnosed before the next condition developed.
Covariates
We adjusted for maternal age and reproductive history, measured by the number of pregnancies prior to each transition. We also adjusted for socioeconomic status using the patient-level IMD9; ethnicity, categorised as White, Asian, Black, Mixed, Other and Unknown12 and region of residence in England (Northeast, Northwest, Yorkshire and The Humber, East Midlands, West Midlands, East of England, London, Southeast and Southwest). In addition, we accounted for the year of preceding LTC diagnosis, grouped into 5-year epochs (<2003, 2003–2007, 2008–2012, 2013–2017 and 2018–2022).
Statistical analysis
Descriptive statistics included frequencies and percentages for categorical variables and means with SD or medians with IQRs for numerical variables. Since participants had varying lengths of follow-up and because some may not experience the event of interest, we used survival analysis methods. These techniques appropriately account for unequal follow-up time and right censoring, allowing for unbiased estimation of the time to transition between LTCs. An ‘event’ was defined as the occurrence of an additional LTC. Women who did not experience the event of interest by the end of follow-up were censored. The entry point at each transition was the date of diagnosis of the preceding LTC. The endpoint for each transition phase was the earliest of the following: the date an additional LTC was diagnosed, the date of death or the last date of data submission by GP practices to the CPRD.
We estimated the median survival time (MST) in years by ethnicity, socioeconomic status and region of residence, as well as the distribution of cumulative failure (ie, acquisition of an additional LTC) from the first LTC to the second LTC, using Kaplan–Meier curves stratified by cohort year of study entry—before 2013 and 2013 onwards. With a violation of the proportional hazard assumption, we used parametric survival analysis to assess factors associated with the time to accrual of additional LTCs (from 1 to 2, 2 to 3, 3 to 4 and 4 to 5+). We used parametric survival methods because of their flexibility in modelling varying survival times, allowing for risks that may increase or decrease over time.13 We performed model diagnostics across candidate parametric models, including Weibull, exponential and log-normal distributions, by comparing Akaike information criterion (AIC) and Bayesian information criterion (BIC) values. These comparisons indicated that the Weibull model provided the best fit for the data relative to the other parametric models.
All analyses were conducted using Stata V.18, with a two-sided significance-level set at 0.05.
Additional analysis
We assessed the timing of transitions from individual baseline conditions with a prevalence greater than 1% to the next LTC diagnosis. Additionally, we examined the timing of transitions for selected baseline conditions known to potentially affect pregnancy outcomes.14 Furthermore, we conducted stratified analyses based on maternal age at study entry (ie, the age at first LTC diagnosis), grouped as 15–24 years, 25–34 years and 35–49 years.
Missing data
We summarised the extent of missing data alongside the characteristics of the study population. Multiple imputations were not undertaken because key variables with missing data (ethnicity and deprivation) could not be reliably predicted from the available dataset and exploratory imputations produced misclassification and unstable estimates. Therefore, a complete-case approach was applied.
Patient and public involvement
Input from patients and the public informed the development and conduct of the study, ensuring its relevance and appropriateness. They will also be engaged in disseminating the results after publication.
Results
At study entry, the mean age was 26.3±7.5 years, and the median number of pregnancies was 0 (IQR: 0–1), increasing gradually with each LTC transition (table 1). Most participants were White, with approximately 9% identifying as Asian and 6% as Black. Of the 2 160 157 women with a single LTC, 65% progressed to develop two or more LTCs.
Table 1. Characteristics of the study population.
| At study entry * | LTC transitions † | ||||
|---|---|---|---|---|---|
| 1 to 2 | 2 to 3 | 3 to 4 | 4 to 5 | ||
| TOTAL | 2 160 157 (100.00) | 1 414 543 (65.48) | 900 621 (41.69) | 555 218 (25.70) | 333 311 (15.43) |
| Characteristics | |||||
| Age (years)‡ | 26.32±7.47 | 29.92±7.58 | 32.24±7.57 | 34.07±7.45 | 35.61±7.26 |
| Number of pregnancies before transition§ | 0 (0,1) | 1 (0,1) | 1 (0,2) | 1 (0,2) | 1 (1,2) |
| Deprivation | |||||
| 1 (Least deprived) | 368 783 (17.07) | 247 516 (67.12) | 156 539 (42.45) | 94 116 (25.52) | 54 589 (14.80) |
| 2 | 367 998 (17.04) | 245 500 (66.71) | 156 266 (42.46) | 95 226 (25.88) | 56 024 (15.22) |
| 3 | 391 594 (18.13) | 254 266 (64.93) | 160 280 (40.93) | 97 554 (24.91) | 57 609 (14.71) |
| 4 | 452 530 (20.95) | 290 328 (64.16) | 183 496 (40.55) | 113 258 (25.03) | 68 696 (15.18) |
| 5 (Most deprived) | 455 973 (21.11) | 303 482 (66.56) | 199 946 (43.85) | 129 009 (28.29) | 81 075 (17.78) |
| Missing | 123 279 (5.71) | 73 451 (59.58) | 44 094 (35.77) | 26 055 (21.13) | 15 318 (12.43) |
| Ethnicity | |||||
| White | 1 422 528 (65.85) | 983 159 (69.11) | 649 641 (45.67) | 410 960 (28.89) | 250 953 (17.64) |
| Asian | 186 563 (8.64) | 115 121 (61.71) | 69 971 (37.51) | 42 024 (22.53) | 25 019 (13.41) |
| Black | 124 012 (5.74) | 74 741 (60.27) | 44 162 (35.61) | 25 582 (20.63) | 14 678 (11.84) |
| Mixed | 31 633 (1.46) | 20 012 (63.26) | 12 303 (38.89) | 7414 (23.44) | 4337 (13.71) |
| Other | 64 333 (2.98) | 36 666 (56.99) | 20 557 (31.95) | 11 531 (17.92) | 6447 (10.02) |
| Unknown | 56 581 (2.62) | 29 582 (52.28) | 15 093 (26.68) | 7414 (13.10) | 3483 (6.16) |
| Missing | 274 507 (12.71) | 155 262 (56.56) | 88 894 (32.38) | 50 293 (18.32) | 28 394 (10.34) |
| Region | |||||
| Northeast | 58 327 (2.70) | 41 347 (70.89) | 27 943 (47.91) | 17 986 (30.84) | 11 116 (19.06) |
| Northwest | 341 314 (15.80) | 240 350 (70.42) | 163 210 (47.82) | 106 267 (31.13) | 66 923 (19.61) |
| Yorkshire and Humber | 69 909 (3.24) | 46 531 (66.56) | 29 836 (42.68) | 18 515 (26.48) | 10 941 (15.65) |
| East Midlands | 52 630 (2.44) | 34 525 (65.60) | 22 168 (42.12) | 13 716 (26.06) | 8337 (15.84) |
| West Midlands | 312 001 (14.44) | 211 318 (67.73) | 138 291 (44.32) | 86 710 (27.79) | 52 588 (16.86) |
| East of England | 112 762 (5.22) | 72 809 (64.57) | 45 169 (40.06) | 27 437 (24.33) | 16 160 (14.33) |
| London | 535 372 (24.78) | 315 620 (58.95) | 184 068 (34.38) | 106 092 (19.82) | 60 477 (11.30) |
| Southeast | 438 405 (20.30) | 289 470 (66.03) | 184 528 (42.09) | 113 224 (25.83) | 67 583 (15.42) |
| Southwest | 239 437 (11.08) | 162 573 (67.90) | 105 408 (44.02) | 65 271 (27.26) | 39 186 (16.37) |
| Cohort entry (first LTC diagnosis) | |||||
| < 2003 | 635 775 (29.43) | 525 168 (82.60) | 392 918 (61.80) | 275 218 (43.29) | 182 720 (28.74) |
| 2003–2007 | 542 350 (25.11) | 369 104 (68.06) | 237 407 (43.77) | 144 494 (26.64) | 84 108 (15.51) |
| 2008–2012 | 483 099 (22.36) | 294 559 (60.97) | 169 828 (35.15) | 92 437 (19.13) | 48 380 (10.01) |
| 2013–2017 | 342 937 (15.88) | 178 251 (51.98) | 85 811 (25.02) | 38 714 (11.29) | 16 718 (4.87) |
| 2018–2022 | 155 996 (7.22) | 47 461 (30.42) | 14 657 (9.40) | 4355 (2.79) | 1385 (0.89) |
Data presented as: n (%)—frequency (per cent).
Column percentage.
Row percentage of the population at study entry.
Mean±SD deviation.
Median (IQR).
LTC, long-term condition.
The median time to develop additional LTCs shortened progressively, from 5.58 years for the transition from 1 to 2 LTCs to 4.45 years from 2 to 3, 3.89 years from 3 to 4 and 3.68 years from 4 to 5 LTCs (table 2). Consistently, women from the most deprived backgrounds had the shortest median transition times for acquiring additional LTCs. Figure 1 shows the 10-year cumulative probability of transitioning from the first LTC to the second LTC. The log-rank test indicated a significant difference in the time to progression between cohorts (χ² = 4260.55, p<0.001), with the Kaplan–Meier curves demonstrating faster progression in the contemporary cohort compared with the earlier cohort.
Table 2. MST of accruing additional LTC*.
| Variables | MST to accrual of LTCs (in years) | |||
|---|---|---|---|---|
| 1–2 | 2–3 | 3–4 | 4–5 | |
| All | 5.58 (1.99, 12.59) | 4.45 (1.57, 10.48) | 3.89 (1.42, 8.52) | 3.68 (1.30, 8.62) |
| Deprivation | ||||
| 1 (Least deprived) | 6.24 (2.28,13.79) | 5.02 (1.82,11.61) | 4.43 (1.65,9.54) | 4.20 (1.53,9.74) |
| 2 | 5.95 (2.14,13.05) | 4.74 (1.70,10.99) | 4.18 (1.55,9.06) | 3.96 (1.41,9.21) |
| 3 | 5.71 (2.04,12.71) | 4.54 (1.63,10.59) | 3.98 (1.48,8.72) | 3.81 (1.36,8.83) |
| 4 | 5.28 (1.91,11.86) | 4.26 (1.50,9.93) | 3.70 (1.34,8.19) | 3.47 (1.23,8.06) |
| 5 (Most deprived) | 4.79 (1.69, 10.71) | 3.81 (1.34,8.88) | 3.36 (1.21,7.40) | 3.21 (1.13,7.47) |
| Ethnicity | ||||
| White | 5.48 (1.99, 12.18) | 4.39 (1.57,10.17) | 3.92 (1.44,8.56) | 3.70 (1.32,8.58) |
| Asian | 5.18 (1.80,12.05) | 4.20 (1.48,10.06) | 3.50 (1.27,7.71) | 3.29 (1.16,7.84) |
| Black | 5.18 (1.83,11.60) | 4.27 (1.50,10.02) | 3.64 (1.32,8.00) | 3.46 (1.23,7.95) |
| Mixed | 5.16 (1.85,11.56) | 4.26 (1.46,10.07) | 3.71 (1.32,7.83) | 3.48 (1.24,8.22) |
| Other | 5.40 (1.83,12.70) | 4.39 (1.49,10.92) | 3.54 (1.27,7.83) | 3.46 (1.19,8.18) |
| Unknown | 7.62 (2.68,17.56) | 6.36 (2.16,15.29) | 5.30 (1.88,11.92) | 5.66 (1.88,13.76) |
| Region | ||||
| Northeast | 5.19 (1.91,11.43) | 4.22 (1.50,9.81) | 3.85 (1.40,8.42) | 3.66 (1.31,8.57) |
| Northwest | 5.03 (1.82,11.16) | 4.02 (1.45,9.30) | 3.65 (1.34,8.04) | 3.45 (1.22,7.84) |
| Yorkshire and The Humber | 5.75 (2.06,13.59) | 4.71 (1.64,11.62) | 4.03 (1.47,8.89) | 3.92 (1.39,10.12) |
| East Midlands | 5.70 (2.02,14.05) | 4.57 (1.59, 12.11) | 3.87 (1.41,8.55) | 3.80 (1.34,9.88) |
| West Midlands | 5.38 (1.95,11.90) | 4.28 (1.54,9.87) | 3.83 (1.40,8.36) | 3.60 (1.27,8.26) |
| East of England | 6.43 (2.29,15.59) | 5.07 (1.79,13.24) | 4.17 (1.55,9.26) | 3.97 (1.42,10.60) |
| London | 5.67 (1.99,12.82) | 4.50 (1.57,10.54) | 3.77 (1.36,8.30) | 3.55 (1.25,8.28) |
| Southeast | 5.86 (2.08,13.21) | 4.64 (1.65,10.95) | 4.04 (1.49,8.90) | 3.84 (1.36,8.91) |
| Southwest | 5.66 (2.00,12.75) | 4.62 (1.62,10.90) | 4.08 (1.47,8.87) | 3.90 (1.37,9.31) |
Data presented as median (IQR).
LTC, long-term condition; MST, median survival time.
Figure 1. 10-year cumulative transition from first LTC to second LTC by cohort. LTC, long-term condition.
Ratios of transition time between LTCs
Online supplemental table S2 and table 3 summarise the unadjusted and adjusted time ratios for the transition between LTCs, respectively. There was a progressive reduction in the time to accrue additional LTCs with decreasing socioeconomic status (table 3). After adjusting for other covariates, women from the most deprived backgrounds acquired a second LTC 21% faster (TR 0.79, 95% CI 0.79 to 0.80) compared with those from the least deprived backgrounds. Although Asian women accrued an additional LTC more slowly during the first transition, they demonstrated faster acquisition of a fourth (TR 0.75, 95% CI 0.74 to 0.75) and fifth LTC (TR 0.93, 95% CI 0.91 to 0.94) than White women. Compared with women in the Southwest region, those in the Northwest, West Midlands and Southeast consistently developed additional LTCs in a significantly shorter time. Accrual of a third, fourth and fifth or more LTCs was also significantly faster in London than in the Southwest.
Table 3. Adjusted time ratios of accrual of LTCs*.
| Variable | LTC transitions † | |||
|---|---|---|---|---|
| 1–2 | 2–3 | 3–4 | 4–5 | |
| Deprivation | ||||
| 1=Least deprived (Ref) | ||||
| 2 | 0.95 (0.95 to 0.96) | 0.95 (0.94 to 0.96) | 0.97 (0.96 to 0.98) | 0.94 (0.93 to 0.95) |
| 3 | 0.92 (0.91 to 0.93) | 0.91 (0.90 to 0.92) | 0.96 (0.95 to 0.97) | 0.91 (0.89 to 0.92) |
| 4 | 0.86 (0.85 to 0.86) | 0.85 (0.85 to 0.86) | 0.93 (0.92 to 0.94) | 0.84 (0.83 to 0.85) |
| 5=Most deprived | 0.79 (0.79 to 0.80) | 0.78 (0.77 to 0.79) | 0.88 (0.88 to 0.89) | 0.78 (0.77 to 0.79) |
| Ethnicity | ||||
| White (Ref) | ||||
| Asian | 1.02 (1.01 to 1.03) | 0.99 (0.98 to 1.00) | 0.75 (0.74 to 0.75) | 0.93 (0.91 to 0.94) |
| Black | 1.00 (0.99 to 1.01) | 1.01 (0.99 to 1.02) | 0.78 (0.77 to 0.79) | 0.97 (0.95 to 1.00) |
| Mixed | 0.99 (0.97 to 1.00) | 1.00 (0.97 to 1.02) | 0.90 (0.88 to 0.93) | 0.97 (0.93 to 1.00) |
| Other | 1.06 (1.05 to 1.07) | 1.03 (1.02 to 1.05) | 0.74 (0.73 to 0.75) | 0.96 (0.93 to 0.99) |
| Unknown | 1.42 (1.40 to 1.44) | 1.43 (1.40 to 1.45) | 1.11 (1.09 to 1.14) | 1.49 (1.43 to 1.55) |
| Region | ||||
| Southwest (Ref) | ||||
| Northeast | 0.95 (0.94 to 0.96) | 0.95 (0.93 to 0.97) | 0.94 (0.92 to 0.96) | 0.98 (0.96 to 1.01) |
| Northwest | 0.90 (0.89 to 0.91) | 0.89 (0.88 to 0.90) | 0.88 (0.87 to 0.89) | 0.89 (0.88 to 0.91) |
| Yorkshire and The Humber | 1.04 (1.02 to 1.05) | 1.03 (1.02 to 1.05) | 0.99 (0.97 to 1.00) | 1.05 (1.02 to 1.08) |
| East Midlands | 1.02 (1.00 to 1.03) | 1.00 (0.99 to 1.02) | 0.97 (0.95 to 0.99) | 0.99 (0.96 to 1.02) |
| West Midlands | 0.94 (0.94 to 0.95) | 0.92 (0.91 to 0.93) | 0.90 (0.89 to 0.91) | 0.92 (0.90 to 0.93) |
| East of England | 1.08 (1.07 to 1.09) | 1.06 (1.05 to 1.08) | 0.97 (0.96 to 0.99) | 1.01 (0.98 to 1.03) |
| London | 1.02 (1.01 to 1.03) | 0.96 (0.95 to 0.97) | 0.88 (0.87 to 0.89) | 0.92 (0.90 to 0.94) |
| Southeast | 0.97 (0.97 to 0.98) | 0.94 (0.93 to 0.95) | 0.92 (0.91 to 0.93) | 0.92 (0.91 to 0.94) |
Adjusted for age at diagnosis, year of preceding diagnosis and the number of pregnancies before each transition.
Data presented as TR (95% CI).
LTCs, long-term conditions.
Stratified analyses by age at first LTC diagnosis were largely consistent with the main findings. However, when women are first diagnosed with an LTC later in the reproductive years (35 years or older) and are from the most deprived backgrounds, they experience a more rapid transition to a second condition (TR 0.72, 95% CI 0.71 to 0.74) compared with women of the same age group from the least deprived backgrounds (table 4). Similarly, older women who were first diagnosed with an LTC and were from ethnic minority groups accrued a second LTC more quickly than White women. Faster acquisition of MLTCs was also observed in the North, West Midlands and Southeast regions, regardless of the age at first LTC diagnosis, compared with the Southwest region.
Table 4. Time ratios for transition from single to MLTCs stratified by age groups at first diagnosis*.
| Variable | Age (in years) † | ||
|---|---|---|---|
| 15–24 | 25–34 | 35–49 | |
| Deprivation | |||
| 1=Least deprived (Ref) | |||
| 2 | 0.95 (0.95 to 0.96) | 0.96 (0.95 to 0.97) | 0.93 (0.91 to 0.95) |
| 3 | 0.93 (0.92 to 0.94) | 0.94 (0.93 to 0.95) | 0.89 (0.87 to 0.91) |
| 4 | 0.87 (0.86 to 0.88) | 0.88 (0.87 to 0.89) | 0.81 (0.79 to 0.83) |
| 5=Most deprived | 0.82 (0.81 to 0.82) | 0.83 (0.82 to 0.84) | 0.72 (0.71 to 0.74) |
| Ethnicity | |||
| White (Ref) | |||
| Asian | 1.10 (1.09 to 1.11) | 0.98 (0.97 to 0.99) | 0.87 (0.85 to 0.89) |
| Black | 1.11 (1.09 to 1.12) | 0.98 (0.97 to 0.99) | 0.88 (0.87 to 0.90) |
| Mixed | 1.01 (0.98 to 1.03) | 0.97 (0.94 to 1.00) | 0.94 (0.90 to 0.99) |
| Other | 1.17 (1.15 to 1.20) | 1.01 (0.99 to 1.03) | 0.94 (0.91 to 0.97) |
| Unknown | 1.46 (1.43 to 1.49) | 1.41 (1.38 to 1.44) | 1.37 (1.32 to 1.42) |
| Region | |||
| Southwest (Ref) | |||
| Northeast | 0.96 (0.95 to 0.98) | 0.94 (0.92 to 0.97) | 0.91 (0.87 to 0.95) |
| Northwest | 0.93 (0.92 to 0.94) | 0.88 (0.86 to 0.89) | 0.84 (0.82 to 0.86) |
| Yorkshire and The Humber | 1.04 (1.02 to 1.06) | 1.05 (1.03 to 1.07) | 1.02 (0.97 to 1.06) |
| East Midlands | 1.02 (1.00 to 1.04) | 1.03 (1.01 to 1.06) | 0.95 (0.90 to 1.00) |
| West Midlands | 0.98 (0.97 to 0.99) | 0.92 (0.90 to 0.93) | 0.87 (0.85 to 0.90) |
| East of England | 1.10 (1.08 to 1.11) | 1.08 (1.06 to 1.10) | 0.97 (0.94 to 1.00) |
| London | 1.07 (1.06 to 1.08) | 0.97 (0.96 to 0.99) | 0.90 (0.87 to 0.92) |
| Southeast | 0.99 (0.98 to 1.00) | 0.97 (0.95 to 0.98) | 0.91 (0.89 to 0.94) |
Adjusted for age at diagnosis, year of preceding diagnosis and the number of pregnancies before each transition.
Data presented as TR (95% CI).
MLTCs, multiple long-term conditions; TR, time ratio.
Transition time based on specific initial conditions
Table 5 presents the median transition time to develop next LTCs from initial conditions that were either prevalent (≥1%) or known to negatively affect pregnancy outcomes. Women with an initial diagnosis of cardiomyopathy had the fastest transition (in years) to a subsequent LTC (MST 3.89; 95% CI 0.99 to 11.78), followed by those with coronary heart disease (MST 4.21; 95% CI 1.30 to 11.34), diabetes (MST 4.25; 95% CI 1.39 to 9.20) and anxiety (MST 4.29; 95% CI 1.46 to 10.12).
Table 5. MST to accrual of another LTC based on selected initial diagnosis.
| Recorded as first diagnosis | Prevalence as first diagnosis (%) ¥ | MST to accrual of next LTC (in years) |
|---|---|---|
| Depression | 9.08 | 4.68 (1.64,10.71) |
| Alcohol abuse | 7.83 | 5.97 (2.04,13.79) |
| Allergic rhinitis | 7.48 | 6.13 (2.27,13.72) |
| Eczema | 7.06 | 6.20 (2.31,13.86) |
| Anxiety | 6.26 | 4.29 (1.46,10.12) |
| Migraine | 6.11 | 5.70 (2.12,12.78) |
| Other psychiatric conditions* | 6.10 | 4.99 (1.77,11.33) |
| Asthma | 5.73 | 6.05 (2.29,12.45) |
| Cancers | 5.42 | 6.81 (2.56,15.41) |
| Infertility | 4.59 | 5.67 (1.90,13.10) |
| IBS | 3.69 | 5.26 (1.99,11.26) |
| PCOS | 2.94 | 5.17 (1.85,11.71) |
| Other dermatological conditions† | 2.42 | 5.55 (2.02,12.91) |
| Thyroid disorder | 2.21 | 6.00 (2.15,13.47) |
| Severe mental illness‡ | 2.14 | 4.51 (1.56,10.08) |
| Vertebral disorder | 1.99 | 8.97 (2.78,32.27) |
| Hypertension | 1.80 | 5.44 (1.94,12.13) |
| Psoriasis | 1.54 | 6.78 (2.57,14.92) |
| Chronic headache | 1.51 | 4.32 (1.52,10.18) |
| Peripheral neuropathy | 1.06 | 5.15 (1.82,12.19) |
| Fibroid | 1.04 | 5.28 (1.90,12.35) |
| Coronary heart disease | 0.10 | 4.21 (1.30,11.34) |
| Stroke | 0.16 | 5.78 (1.79,13.33) |
| Cardiomyopathy | 0.02 | 3.89 (0.99,11.78) |
| Valvular heart disease | 0.16 | 6.68 (2.55,13.89) |
| Atrial fibrillation | 0.02 | 4.98 (1.57,11.21) |
| Epilepsy | 0.59 | 7.64 (2.82,15.45) |
| Venous thromboembolism | 0.58 | 6.12 (2.29,13.07) |
| Diabetes mellitus | 0.65 | 4.25 (1.39,9.20) |
| SLE | 0.07 | 5.53 (1.77,11.27) |
Other psychiatric conditions—obsessive compulsive disorder, self-harm, personality disorder and dissociative disorder.
Other dermatological conditions—seborrheic dermatitis, rosacea, hidradenitis suppurativa and lichen planus.
Severe mental illness—bipolar affective disorder, schizophrenia and psychosis.
IBS, irritable bowel syndrome; LTC, long-term condition; MST, median survival time; PCOS, polycystic ovary syndrome; SLE, systemic lupus erythematosus.
For most conditions, women in the most deprived quintiles developed additional conditions more rapidly, progressing faster to MLTCs than those from the least deprived backgrounds. Exceptions included a few low-prevalence conditions, such as stroke, cardiomyopathy, valvular heart disease, atrial fibrillation and systemic lupus erythematosus, for which no statistically significant differences in transition time based on socioeconomic status were observed (online supplemental table S3).
The transition time for White and minority ethnic groups was largely comparable when we considered specific initial diagnosis. However, Black women whose initial diagnosis was depression, alcohol misuse or infertility experienced shorter transition times to the next LTC than White women. Similarly, Asian women with an initial diagnosis of polycystic ovarian syndrome, thyroid disorder, infertility, hypertension or dermatological conditions (such as seborrhoeic dermatitis and rosacea) also progressed more quickly to a subsequent LTC (online supplemental table S3).
Discussion
Summary of findings
This study offers unique insight into the transition time between LTCs among childbearing women in England, highlighting variation based on socioeconomic status, ethnicity and region of residence. The median transition time between LTCs progressively shortened, indicating a cumulative acceleration in the burden of MLTCs. The contemporary cohort transitioned to MLTCs more quickly. Consistently, socioeconomic deprivation was associated with accelerated LTC accumulation. In addition, a regional divide in LTC accumulation was evident, with women living in the northern regions of England, the West Midlands, the Southeast and London generally exhibiting significantly faster accumulation of additional LTCs compared with those in the Southwest. Although transition times between LTCs were largely similar across ethnic groups, for some individual conditions, ethnicity was associated with a more rapid progression to MLTCs. When baseline condition-specific trajectories were examined, cardiometabolic conditions, particularly cardiomyopathy, coronary heart disease and diabetes, showed the fastest progression to MLTC.
Comparison with existing literature
The cumulative acceleration in the burden of MLTCs over time aligns with prior literature, emphasising the worsening of health as additional chronic conditions develop.15 16 Our findings, indicating that socioeconomic status consistently influences LTC accrual, are consistent with Ribe et al, who observed faster disease accumulation among middle-aged and older adults from the most disadvantaged backgrounds compared with their least deprived peers.17 This reflects longstanding evidence of the social gradient in health and highlights persistent structural inequalities in the prevention, diagnosis and management of chronic conditions.18 19
A few studies have shown that minority ethnic groups, particularly those of South Asian descent, may be at a higher risk of MLTCs in the UK.20 21 We found that, in later transitions, Asian women developed additional LTCs at a faster rate—suggesting a combination of genetic susceptibility and other structural and behavioural factors, including diet and health-seeking behaviour. Regional disparities in the transition time between LTCs highlight geographic patterns of health inequalities, potentially influenced by variations in environmental exposures, access to services, health behaviours and regional policy decisions.22
The finding of faster transition to MLTCs in the contemporary cohort aligns with a Canadian study that examined multimorbidity prevalence across birth cohorts from 1925 to 1974, which showed higher prevalence of MLTCs in more recent cohorts.23 The prevalence of MLTCs—particularly complex multimorbidity, where an individual has four or more conditions—is projected to increase due to improved diagnostic practices and unhealthy dietary and lifestyle behaviours.24 25 This trend may help explain the differences in progression observed between the earlier and contemporary cohorts in our study. The faster accumulation of additional LTC following cardiometabolic disease is consistent with evidence from previous longitudinal and multimorbidity studies.26 27 Anxiety has also been shown to accelerate the accumulation of physical diseases.28 Our study adds novel insight by quantifying transition times by baseline condition in childbearing women using large-scale UK primary care data.
Implications for practice
Our findings have important implications for clinical practice and public health policy. The observed acceleration in the burden of MLTCs, particularly in the contemporary cohort, may reflect increasing physiological vulnerability, contemporary lifestyle changes, advancements in medical diagnosis or growing challenges in disease management as conditions accumulate. More proactive interventions would be necessary for women from socioeconomically deprived backgrounds diagnosed with an initial LTC to prevent or delay progression to MLTCs. Early intervention strategies, emphasising lifestyle changes, mental health support and cardiovascular risk reduction, should be integrated as priorities within routine reproductive and general healthcare services.
It is possible that the variation in the transition time between LTCs by ethnicity and socioeconomic status stems from complex interactions among biological susceptibility, access to healthcare and broader social determinants of health. Policies must address the unequal burden of MLTCs among ethnic minorities and disadvantaged groups through culturally competent care, targeted screening and better access to mental health and chronic disease services. Investment in integrated, community-based care could be particularly valuable for women with multiple conditions. Additionally, clinical guidelines should specifically account for MLTCs risk profiles when managing common initial conditions, such as depression.
The predominance of short accrual times across multiple baseline conditions in regions, such as the Northwest and West Midlands, highlights an urgent need for targeted health-promotion interventions in these areas. Regional policymakers and healthcare planners should prioritise directing resources to high-burden areas.
The finding that cardiometabolic conditions exhibited the fastest progression to MLTCs likely reflects shared pathophysiological mechanisms, including chronic inflammation, endothelial dysfunction and metabolic dysregulation. Additionally, the presence of anxiety among the fastest trajectories underscores the role of mental health as an early and influential component of MLTC development, rather than merely a downstream consequence. Importantly, the observed condition-specific trajectories reflect real-world clinical pathways in primary care, rather than patterns derived from selected trial populations. Moreover, baseline condition-specific trajectories may facilitate earlier identification of women at high risk of rapid MLTC accumulation in primary care. In particular, women presenting with cardiomyopathy, coronary heart disease, diabetes or anxiety represent high-risk entry points into multimorbidity, and these conditions could serve as early flags for proactive MLTC prevention.
Strengths and limitations
The study’s large sample size and longitudinal design enhance the robustness and reliability of the findings. However, some limitations are acknowledged. Reliance on routine healthcare data may lead to under-reporting or misclassification of LTCs. The absence of statistical significance for the less prevalent conditions is likely due to imprecision arising from small case numbers rather than an absence of a true difference. In the condition-specific stratified analyses, we acknowledge the wide CIs, particularly for cardiomyopathy, which likely reflect smaller sample sizes; however, the ordering of trajectories remains clinically plausible and coherent. Further stratification by disease severity was not feasible due to constraints inherent in routinely collected primary care data, where consistent and validated severity measures are not uniformly available across conditions. However, the approach of stratifying analyses by baseline condition-specific trajectories remains relevant for practical application, as it reflects real-world clinical pathways captured in primary care and can inform population-level risk stratification, service planning and early intervention strategies. Furthermore, while clear associations with deprivation and, in some instances, ethnicity were observed, unmeasured confounders, such as health literacy, care-seeking behaviour and cultural influences, may partially account for the observed disparities in the time to accrue additional LTC.
Conclusion
This study offers evidence on the transition time between LTCs among childbearing women in England, driven by socioeconomic deprivation, ethnicity and residence. There is an urgent need to embed MLTC prevention within reproductive health services, tailor interventions for high-risk groups and tackle the structural factors underpinning health inequalities across the life course.
Future research should investigate the influence of health behaviours and environmental factors on the burden of MLTCs among childbearing women.
Supplementary material
Footnotes
Funding: This research was funded by the Trailblazers Early-Career Researcher and Doctoral Studentship Partnering Scheme at Coventry University (Award No. 13771‑93) awarded to AA. The funder was not involved in the study’s design, data collection, analysis, interpretation or the preparation of this report.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Data availability free text: The study draws on data provided by the Clinical Practice Research Datalink (CPRD), accessed under a licence agreement with the UK Medicines and Healthcare products Regulatory Agency. Due to licensing restrictions, the dataset cannot be shared publicly. Researchers interested in similar data can apply directly to CPRD, subject to approval by the Independent Scientific Advisory Committee. The data are derived from patient information recorded by the NHS as a part of routine clinical practice. All interpretations and conclusions are those of the authors and do not necessarily reflect the views of the data providers.
Patient and public involvement: Patients and/or the public were involved in the design, or conduct, or reporting or dissemination plans of this research. Refer to the Methods section for further details.
Author note: The guarantor affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.
Data availability statement
Data may be obtained from a third party and are not publicly available.
References
- 1.Knight M, Bunch K, Tuffnell D, et al. Saving lives, improving mothers’ care – lessons learned to inform maternity care from the UK and Ireland confidential enquiries into maternal deaths and morbidity 2017–19. Oxford: National Perinatal Epidemiology Unit, University of Oxford; 2021. [Google Scholar]
- 2.Admon LK, Winkelman TNA, Moniz MH, et al. Disparities in Chronic Conditions Among Women Hospitalized for Delivery in the United States, 2005-2014. Obstet Gynecol. 2017;130:1319–26. doi: 10.1097/AOG.0000000000002357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stanhope KK, Worrell N, Jamieson DJ, et al. Double, Triple, and Quadruple Jeopardy: Entering Pregnancy With Two or More Multimorbid Diagnoses and Increased Risk of Severe Maternal Morbidity and Postpartum Readmission. Womens Health Issues. 2022;32:607–14. doi: 10.1016/j.whi.2022.06.005. [DOI] [PubMed] [Google Scholar]
- 4.Bestman PL, Kolleh EM, Moeng E, et al. Association between multimorbidity of pregnancy and adverse birth outcomes: A systemic review and meta-analysis. Prev Med. 2024;180:107872. doi: 10.1016/j.ypmed.2024.107872. [DOI] [PubMed] [Google Scholar]
- 5.NIHR Evidence . Multiple long-term conditions (multimorbidity) and inequality: addressing the challenge: insights from research. 2023. [Google Scholar]
- 6.World Health Organization . Reproductive health indicators: guidelines for their generation, interpretation and analysis for global monitoring. Geneva: 2006. [Google Scholar]
- 7.Clinical Practice Research Datalink CPRD GOLD November 2024 (version 2024.11.001) 2024. [5-Dec-2024]. Available. Accessed. [DOI]
- 8.Clinical Practice Research Datalink CPRD Aurum September 2024 (version 2024.09.001) 2024. [5-Dec-2024]. Available. Accessed. [DOI]
- 9.Clinical Practice Research Datalink Small area level data based on patient postcode (version 3.2: documentation and data dictionary (set 22/January 2022)) 2022. [6-Dec-2024]. https://www.cprd.com/sites/default/files/2022-02/Documentation_SmallAreaData_Patient_set22_v3.2.pdf Available. Accessed.
- 10.Lee SI, Azcoaga-Lorenzo A, Agrawal U, et al. Epidemiology of pre-existing multimorbidity in pregnant women in the UK in 2018: a population-based cross-sectional study. BMC Pregnancy Childbirth. 2022;22:120. doi: 10.1186/s12884-022-04442-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.MuMPreDiCT Exposure multimorbidity clincial codes: mumpredict. 2024. https://github.com/mumpredict/Exposure-multimorbidity-clinical-codes- Available.
- 12.Office for National Statistics Ethnic group classifications: census 2021. 2021. https://www.ons.gov.uk/census/census2021dictionary/variablesbytopic/ethnicgroupnationalidentitylanguageandreligionvariablescensus2021/ethnicgroup/classifications Available.
- 13.Royston P, Lambert PC. Flexible parametric survival analysis using Stata: beyond the Cox model. College Station, TX: Stata Press; 2011. [Google Scholar]
- 14.Felker A, Patel R, Kotnis R, et al. Saving lives, improving mothers’ care: lessons learned to inform maternity care from the UK and Ireland confidential enquiries into maternal deaths and morbidity 2021–23. Oxford: National Perinatal Epidemiology Unit: University of Oxford; 2025. [Google Scholar]
- 15.Barnett K, Mercer SW, Norbury M, et al. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380:37–43. doi: 10.1016/S0140-6736(12)60240-2. [DOI] [PubMed] [Google Scholar]
- 16.Singer L, Green M, Rowe F, et al. Trends in multimorbidity, complex multimorbidity and multiple functional limitations in the ageing population of England, 2002-2015. J Comorb . 2019;9:2235042X19872030. doi: 10.1177/2235042X19872030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ribe E, Cezard GI, Marshall A, et al. Younger but sicker? Cohort trends in disease accumulation among middle-aged and older adults in Scotland using health-linked data from the Scottish Longitudinal Study. Eur J Public Health. 2024;34:696–703. doi: 10.1093/eurpub/ckae062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Canning D, Bowser D. Investing in health to improve the wellbeing of the disadvantaged: reversing the argument of Fair Society, Healthy Lives (the Marmot Review) Soc Sci Med. 2010;71:1223–6. doi: 10.1016/j.socscimed.2010.07.009. [DOI] [PubMed] [Google Scholar]
- 19.Álvarez-Gálvez J, Ortega-Martín E, Carretero-Bravo J, et al. Social determinants of multimorbidity patterns: A systematic review. Front Public Health. 2023;11:1081518. doi: 10.3389/fpubh.2023.1081518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Eto F, Samuel M, Henkin R, et al. Ethnic differences in early onset multimorbidity and associations with health service use, long-term prescribing, years of life lost, and mortality: A cross-sectional study using clustering in the UK Clinical Practice Research Datalink. PLoS Med. 2023;20:e1004300. doi: 10.1371/journal.pmed.1004300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hayanga B, Stafford M, Bécares L. Ethnic inequalities in multiple long-term health conditions in the United Kingdom: a systematic review and narrative synthesis. BMC Public Health. 2023;23:178. doi: 10.1186/s12889-022-14940-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Office for National Statistics . Health inequalities. 2021. [2-Apr-2024]. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthinequalities Available. accessed. [Google Scholar]
- 23.Canizares M, Hogg-Johnson S, Gignac MAM, et al. Increasing Trajectories of Multimorbidity Over Time: Birth Cohort Differences and the Role of Changes in Obesity and Income. J Gerontol B Psychol Sci Soc Sci. 2018;73:1303–14. doi: 10.1093/geronb/gbx004. [DOI] [PubMed] [Google Scholar]
- 24.McGrail K, Lavergne R, Lewis S. The chronic disease explosion: artificial bang or empirical whimper? BMJ. 2016;352:i1312. doi: 10.1136/bmj.i1312. [DOI] [PubMed] [Google Scholar]
- 25.Haapanen MJ, Vetrano DL, Mikkola TM, et al. Early growth, stress, and socioeconomic factors as predictors of the rate of multimorbidity accumulation across the life course: a longitudinal birth cohort study. Lancet Healthy Longev. 2024;5:e56–65. doi: 10.1016/S2666-7568(23)00231-3. [DOI] [PubMed] [Google Scholar]
- 26.Ioakeim-Skoufa I, Ledesma-Calvo R, Moreno-Juste A, et al. Charting the Pathways of Cardiometabolic Multimorbidity: A Systematic Review of Clinical Trajectories. J Clin Med. 2025;14:2615. doi: 10.3390/jcm14082615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhang D, Tang X, Shen P, et al. Multimorbidity of cardiometabolic diseases: prevalence and risk for mortality from one million Chinese adults in a longitudinal cohort study. BMJ Open. 2019;9:e024476. doi: 10.1136/bmjopen-2018-024476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bobo WV, Grossardt BR, Virani S, et al. Association of Depression and Anxiety With the Accumulation of Chronic Conditions. JAMA Netw Open. 2022;5:e229817. doi: 10.1001/jamanetworkopen.2022.9817. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data may be obtained from a third party and are not publicly available.