Background
Methods
Search strategy
Study selection
Data extraction
Results
Study selection
Study characteristics
First author, publication year | Years of study | Location of study | Exclusion criteria | Cases | Controls | Age | Endpoint categorya
| Specific endpoint | ||
---|---|---|---|---|---|---|---|---|---|---|
CHD | CV A | CVD | ||||||||
Beard 1984 [23] | 1960–1974 | Rochester, Minnesota, USA | Unmarried women. | 169, hospital admissions | 338 age-matched (3y), seen at same clinic in same year as diagnosis for matched case, no prior CVD diagnosis | < 60 | C | Incident cases of CHD (angina, myocardial infarction, sudden unexpected death) | ||
La Vecchia 1987 [31] | 1983–1986 | Northern Italy | Women with recurrent infarction or uncertain diagnostic criteria. | 202, hospital admissions | 374, under age 55 admitted to same hospital with acute disease other than CVD, malignant, metabolic, hormonal or gynecologic (and not related to smoking/alcohol) | < 55 y, median age 47 (cases) 45 (controls: 24–54) | C | First time myocardial infarction | ||
Talbott 1989 [37] | 1973–1975 | Pennsylvania, USA | Never married, no living relatives, prior history of CVD | 67 cases (death records in one county) | 73 neighborhood controls, alive, within 10 years age match, Caucasian. | 25–64, mean age 54.6 (cases) 43.4 (controls) | M | Sudden cardiac death | ||
Palmer 1992 [34] | 1986–1990 | Massachusetts, USA | Controls: people without phone-listing excluded (27%). | 858, hospital admissions | 858 controls from registry, 5 years age range, same precinct of residence. | 45–69, mean age 60 | E | First time, non-fatal myocardial infarction | ||
Okamoto 2001 [33] | 1992–1997 | Nagoya, Japan | Aged ≥80 or <30 | 124 consecutive cases | 248 age-matched (2 y) controls, one hospital (patients with gastro-intestinal disease) and one community control per case. | 30–79, mean age 60 (cases) and 60.3 (controls) | C | First time subarachnoid hemorrhage | ||
Bertuccio 2007 [24] | 3 studies: 1983–1992 1988–1989 1995–2003 | Italy | None mentioned | 609, in hospital | 1106 (in hospital for acute diseases unrelated to smoking and other recognized myocardial infarction risk factors) | 18–79, median age 56 y (cases) 53 y (controls) | E | First time myocardial infarction |
First author, publication year | Duration of study + years of FU | Location of study | Exclusion criteria | Cohort size | Number of cases (% of cohort) | Age at entry | Endpoint category a
| Specific endpoint | ||
---|---|---|---|---|---|---|---|---|---|---|
CHD | CVA | CVD | ||||||||
Colditz 1987 [26] | En: ‘76 F: until ‘82 FU: 6 y | USA | Diagnoses of CHD at entry, non-married. (Note: only registered nurses, 98% white) | 119,963 women. 700,809 person years | 308 (2.6%) | 30–55 at entry | C | Incident cases non-fatal myocardial infarction or fatal CHD. | ||
Cooper 1999 [27] | En: ‘34-‘39 F: until ‘90–‘91. FU: 51–57 y | Minne-sota, USA | Enrolled white college students. In FU excluded if missing information age CVD event, or missing CVD data in questionnaire | 867. (714 self-administered FU, 153 proxy respondents (family members)) | 45 (35 non-fatal, 10 fatal) (5.2%) | Under 25 at entry, 63–81 at FU | C | Non-fatal and fatal CHD | ||
Otterblad- Olausson 2004 [9] | En:‘85 census. FU: From ‘90-‘95. FU: 5 y | Sweden | Women who had first infant before the age of 30 between 1964 and 1989 are included. Exclusion: women who emigrated or died between census and follow up. | 460,434 women | 151 (total circulatory) (0.05%), no data reported for CHD and cerebrovascular. | Born ‘50 – ‘64 → 21 to 35 at census. | M | M | M | Total circulatory, CHD, cerebrovascular mortality. |
Yang CY 2006 [38] | En: ‘78/‘87 F: until ‘03 FU: 16–25 y | Taiwan | Included women with a first and singleton childbirth between 1978 and 1987. | 1,292,462 women, 27,402,995 person years | 189 (0.01%) | ? | M | Subarachnoid hemorrhage mortality | ||
Henrettab
2007 [30] | En: ‘31/‘41 F: ‘92/‘02a
| USA | First two follow up years excluded. People who didn’t survive until follow up excluded. | 4335 women | Heart disease: 13.2%, stroke: 8.9% (parous women) | Birth cohort ‘31-‘41. Age at FU: 51–61 through 61–71 | E | E | Presence of disease in 1994: heart, stroke. Self-reported. | |
Sakauchi 2007 [36] | En:‘88/‘90 F: until ‘03. FU: 13–15 y | Japan | Previous history of cancer. Missing target question item. | 63,600 women at baseline, 817,669 person years | CHD: 458 (0.7%) Cerebrovascular: 1151 (1.8%) | 40 to 79 at entry. | M | M | CHD mortality, cerebrovascular mortality | |
Yang L 2009 [39] | En: ‘91-‘92 F: until ‘04. FU: avg. 12.9 y | Sweden | Prior CVD. Missing information on OC use or smoking. Baseline self-reported natural menopause or HRT use. | 45,729 | Stroke: 285 (0.6%); Ischemic: 193 (0.4%) Hemorrhagic: 72 (0.2%) | 30–49 | C | Ischemic and Hemorrhagic stroke. | ||
Grundy 2010 [29] | En: ‘35/‘68 F:’ from ‘80 to ‘03. FU: 23 y | Norway | Analysis restricted to ages above 45 (largely completed childbearing for women). Men and women who died before this age were excluded | 744,784 women (7.2 million person years) 785,317 men (7.36 million person years) | 3605 women (0.5%) 12,640 men (1.6%) | Register data born ‘35-‘68. Mortality at 45–68 | M | Death from circulatory diseases | ||
Chang 2011 [25] | En: ‘85 FU: until ‘05 FU: 20 y | Kangwha, Korea | Ever on OC. Missing info on age menarche, MP, BMI, BP, AFB, gravidity or parity. | 3257 women. 48,313 person years | 478 (14.7%) | PostMP, 55 or older at entry | M | M | M | CVD mortality, CHD mortality, stroke mortality |
Gallagher 2011 [28] | En: ‘89-‘91 F: until ‘00 FU: 10/12 y | China | None mentioned | 267,400 women. 2,565,433 person years | CVD: 4349 (1.6%), Stroke: 2776 (0.9%), CHD: 494 (0.2%). | 30–60 at entry | M | M | CHD, ischemic and hemorrhagic stroke mortality. | |
Webb 2011 [10] | En: ‘70 F: until ‘06 FU: max 37 y | England and Wales | Included women estimated to have reached age 13 during ‘70s-‘90s, sampled members of census ‘71, ‘81 and ‘91. | 1,913,595 person years | 133 (0.007%) | 13–49 | M | Heart disease mortality | ||
Merritt 2015 [32] | En: ‘92/‘00 F: Until ‘10. FU: mean 12.9 y | 10 western European countries | DM, myocardial infarction, angina, stroke, cancer. Report of a prevalent disease that could influence important confounders. Missing lifestyle questionnaire, vital status or date of death. Women reported never having menstruated or missing information on all reproductive variables. | 322,972 women | Circulatory: 2404 (0.7%); Cerebrovascular: 808 (0.3%); CHD: 732 (0.2%) | 25–70 at entry | M | M | M | Total circulatory, cerebrovascular, and CHD mortality |
Barclay 2016 [22] | En: ‘32 - ‘60 F: ‘90 - ‘12 FU: 22 y | Sweden | This is a sibling study: ID for both parents needed, no multiple births, no only children, no childless individuals, no variance on either mortality or AFB. | 12,635 women, 27,183 men | 4503 women (35.6% 10,067 men (37.0%) | People enter cohort in ‘90 or after age 45. Ages 45–80 throughout follow-up. | M | Circulatory deaths | ||
Parikh 2016 [35] | En: ‘91 F: until ‘10 FU: median 12 years | USA | Missing reproductive and CHD risk factor information, missing follow up, prevalent or unknown history of CHD. | 72,982 | 4607(6.3%) | Mean age at start of study 63.2 | E | Coronary Heart Disease events |
Study findings
Summarizing study results
Case control studies | |||||||
---|---|---|---|---|---|---|---|
First author, publication year | Age at first birth | Exposure cases: N (%) | Exposure controls N (%) | Point estimate (+CI) a
Base model
| Point estimate (+CI) a
Final model
| Adjusted for or matched on. If underlined, also adjusted in base model. | Notes on study considerations and limitations. |
Beard
1984 [23] | < 20 | Data not provided | Data not provided | 1.9 (0.7–5.6) | D: age; SE: - BM: HT, DM; HB: smoking; R: - | No data on number of individuals per exposure group. Likely number for <20 group is small. | |
20–24 |
1.8 (1.1–3.3)
| ||||||
≥25/n.a. b
| 1 (Ref) | ||||||
La Vecchia
1987 [31] | < 20 | 21 (10.4) | 23 (6.1) |
2.31 (1.1–4.9)
|
c D: age; SE: - BM: - HB: - R: - | Nulliparous group includes young women who did not have children yet. | |
20–24 | 71 (35.1) | 117 (31.3) | 1.47 (0.9–2.4) | ||||
≥ 25 | 75 (37.1) | 111(29.7) | 1.39 (0.8–2.4) | ||||
Nullipara | 35 (17.3) | 123 (32.9) | 1 (ref) | ||||
Talbott [37]
1989
| <20 | 14 (26.9) | 7 (10.3) |
3.4 (1.1–9.9)
| 2.5 (0.5–12.8) | D: age at demise (base model) SE: - BM: - HB: smoking (final model); R: - | No adjustment for age in final model, despite much younger controls. Small sample size. Caucasian-only controls; no race/ethnicity information for cases. |
≥ 20 | 38 (73.1) | 61 (89.7) | |||||
Palmer
1992 [34] | < 18 | 34 (4.5) | 5 (0.7) | 6.8 |
5.5 (1.7–17)
| D: marital status (and age-matched); SE: education, education spouse, occupation; BM: drug-treated HT, elevated serum cholesterol, drug-treated DM, family history MI, BMI; HB: smoking, coffee, alcohol, physical activity; R: conjugated estrogen use, age menarche, parity, menopausal status | Extensive adjustment in final model |
18 | 33 (4.3) | 31 (4.2) | 1.2 | 1.1 (0.6–2.2) | |||
19 | 65 (8.5) | 39 (5.3) | 1.6 | 1.6 (0.9–2.8) | |||
20–24 | 376 (49.4) | 366 (49.4) | 1 | 1 (ref) | |||
25–29 | 177 (23.3) | 224 (30.2) | 0.7 | 0.8 (0.6–1.1) | |||
30–34 | 61 (8.0) | 58 (7.8) | 1.0 | 1.3 (0.8–2.2) | |||
≥ 35 | 15 (2.0) | 18 (2.4) | 0.8 | 1 (0.4–2.9) | |||
<20 | 132 (17.3) | 75 (10.1) | CI not provided |
1.7 (1.1–2.6)
| |||
≥ 20 | 629 (82.7) | 666 (89.9) | 1 (ref) | ||||
Okamoto
2001 [33] | < 27 | 61 (49.0) | 146 (58.8) | 1 (ref) | D: age; SE: educational level; BM: hypertension; HB: smoking; R: - | Cut-off for youngest group <27 | |
≥ 27 | 63 (51.0) | 102 (41.2) | 1.45 (0.9–2.3) | ||||
Bertuccio
2007 [24] | < 20 | 58 (11.4) | 73 (8.6) | 1 (ref) | 1 (ref) | D: age, study; SE: education; BM: BMI, Cholesterol, DM, obesity, HLD, HT; HB: Smoking, coffee, alcohol; R: parity, MP, HRT, family history AMI | Extensive adjustment in final model |
20–24 | 207 (40.7) | 342 (40.1) | 0.72 (0.5–1.1) | 0.90 (0.6–1.4) | |||
25–29 | 172 (33.9) | 319 (37.4) |
0.65 (0.4–0.97)
| 0.92 (0.6–1.5) | |||
≥ 30 | 71 (14.0) | 118 (13.9) | 0.70 (0.4–1.1) | 0.94 (0.6–1.6) |
Cohort Studies | |||||||
---|---|---|---|---|---|---|---|
First author, publication year | Age at first birth | No. in cohort | No. study outcomes N (%) | Point estimate (+CI) a
Base model
| Point estimate (+CI) a
Final model
| Adjusted for or matched on. If underlined, also adjusted in base model. | Notes on study considerations and limitations. |
Colditz
1987 [26] | <19 | 6671 b
| 3 (0.004) | 1.3 (0.4–4.3) |
c D: age; SE: - BM: - HB: - R: - | Very few events in the exposure category of <19. Short follow-up in a relatively young age group. | |
20–22 | 145,708 | 42 (0.029) | 0.9 (0.6–1.2) | ||||
23–25 | 257,170 | 108 (0.042) | 1 (ref) | ||||
26–29 | 151,206 | 82 (0.54) | 1.1 (0.8–1.5) | ||||
≥ 30 | 77,091 | 35 (0.045) | 0.8 (0.5–1.2) | ||||
Cooper
1999 [27] | 15–24 | 7996 b
| 8 (0.100) | 1.29 (0.5–3.0) | D: age; SE: - BM: - HB: - R: - | Number of outcomes very low in all exposure categories | |
25–29 | 17,698 | 14 (0.079) | 1 (ref) | ||||
30–32 | 5341 | 3 (0.056) | 0.71 (0.2–2.5) | ||||
33–43 | 3715 | 8 (0.215) |
2.9 (1.2–6.9)
| ||||
Otterblad- Olausson
2004 (CVD) [9] | <20 | 60,686 | 41 (0.07) |
2.2 (1.5–3.1)
|
1.8 (1.2–2.6)
| D: age; SE: background SES in 1960 (parents), socio-economic position (‘90), family situation (‘90), welfare dependency (‘90) BM: - HB: - R: Parity | Conducted amongst women aged 30–45, follow-up time of 5 years, resulting in low number of events. Some women still having children. |
20–29 | 399,748 | 110 (0.03) | 1 (ref) | 1 (ref) d
| |||
Otterblad- Olausson
2004 (CHD) [9] | <20 | 60,686 | Data not |
2.8 (1.5–5.4)
|
2.2 (1.2–4.3)
| ||
20–29 | 399,748 | provided | 1 (ref) | 1 (ref) d
| |||
Otterblad- Olausson
2004 (CeVD) [9] | <20 | 60,686 | Data not | 1.5 (0.8–2.9) | 1.4 (0.7–2.7) | ||
20–29 | 399,748 | provided | 1 (ref) | 1 (ref) d
| |||
Yang CY 2006
| <26 | 859,942 | 102 (0.012) | 1 (ref) | D: - SE: - BM: - HB: - R: Parity | No adjustment for participant age is reported (while due to study set-up women with older age at first birth were older at follow-up). Cut-off for youngest group <26 | |
26–30 | 372,895 | 70 (0.019) |
1.78 (1.3–2.4)
| ||||
≥ 31 | 59,625 | 17 (0.029) |
2.96 (1.8–5.0)
| ||||
Continuous |
1.10 (1.06–1.1)
|
1.08 (1.04–1.1)
| |||||
Henretta
2007
Heart disease [30] | <20 | 991 | Data not provided |
0.36 (P:<.01)
| D: age, race, US-born, unmarried at first birth, marital status; SE: Father’s education, education, log net worth, log income; BM: - HB: - R: Birth >39, birth interval, parity | No information on the number of outcomes per exposure group; no confidence intervals | |
≥ 20 | 2956 | ||||||
Henretta
2007
Stroke [30] | <20 | 991 | Data not provided | −.03
P-value or CI not provided | |||
≥ 20 | 2956 | ||||||
Sakauchi 2007
CHD [36] | <23 | Data not provided | 121 | 1.09 (0.8–1.5) | D: age, study area; SE: - BM: - HB: - R: - | No denominator information provided | |
23–25 | 116 | 1 (ref) | |||||
≥ 26 | 131 | 1.17 (0.9–1.5) | |||||
Sakauchi
2007
(CeVD) [36] | <23 | Data not provided | 291 | 1.19 (0.99–1.4) | |||
23–25 | 294 | 1 (ref) | |||||
≥ 26 | 326 | 1.07 (0.9–1.3) | |||||
Yang L
2009, Ischemic stroke [39] | <23 | 11,942 | 68 (0.57) | 1.1 (0.7–1.5) | 1.0 (0.6–1.5) | D: age; SE: education; BM: BMI, high blood pressure, DM; HB: alcohol, smoking, physical activity; R: - | Young age at first birth category not compared to lowest risk category. Few cases for hemorrhagic stroke. |
21–25 | 9905 | 49 (0.49) | 1 (ref) | 1 (ref) | |||
≥ 26 | 17,444 | 48 (0.28) |
0.6 (0.4–0.9)
| 0.7 (0.4–1.1) | |||
Yang L | <21 | 11,942 | 28 (0.23) |
2.0 (1–4)
| 1.8 (0.8–4.1) | ||
2009
hemorrhagic stroke
| 21–25 | 9905 | 11 (0.11) | 1 (ref) | 1 (ref) | ||
≥ 26 | 17,444 | 19 (0.11) | 1.1 (0.5–2.3) | 1.2 (0.5–2.6) | |||
Grundy
2010 [29]
Women
| <20 | 862,007 b
| 572 (0.07) |
1.47 (1.3–1.6)
|
1.22 (1.1–1.3)
| D: age, year, region residence, log population size, marital status; SE: level of education; BM: - HB: - R: parity
| |
20–24 | 3,200,462 | 1567 (0.05) | 1 (ref) | 1 (ref) | |||
25–29 | 1,670,417 | 574 (0.03) |
0.67 (0.6–0.7)
|
0.81 (0.7–0.9)
| |||
≥ 30 | 646,983 | 224 (0.03) |
0.60 (0.5–0.7)
|
0.78 (0.7–0.9)
| |||
Grundy
2010
Men [29] | <23 | 1,164,183 b
| 2190 (0.19) |
1.39 (1.3–1.5)
|
1.23 (1.2–1.3)
| ||
23–28 | 3,238,174 | 4804 (0.15) | 1 (ref) | 1 (ref) | |||
29–34 | 1,313,342 | 1772 (0.13) |
0.88 (0.8–0.9)
|
0.93 (0.9–0.99)
| |||
≥ 35 | 473,472 | 571 (0.12) |
0.87 (0.8–0.95)
| 0.93 (0.9–1.01) | |||
Chang
2011
CVD [25] | 15–19 | 844 | 144 (17.1) | 1.00 (ref) | 1.00 (ref) d
| D: age at entry; SE: education, occupation; BM: BMI, HT HB: drinking, smoking R: - For CVD there was a mention of a fourth model in which the authors additionally corrected for reproductive variables and marital status, which did not change the findings. | Extensive adjustment in final model. Few CHD events. |
20–22 | 1646 | 230 (14.0) |
0.74 (0.6–0.9)
|
0.74 (0.6–0.9)
| |||
≥ 23 | 787 | 104 (13.2) |
0.75 (0.6–0.96)
|
0.76 (0.6–0.98)
| |||
Chang
2011
CHD [25] | 15–19 | 844 | 12 (1.42) | 1.00 (ref) | 1.00 (ref) d
| ||
20–22 | 1646 | 24 (1.46) | 0.88 (0.4–1.8) | 0.89 (0.4–1.8) | |||
≥ 23 | 787 | 11 (1.40) | 0.93 (0.4–2.1) | 0.90 (0.4–2.1) | |||
Chang
2011
Stroke [25] | 15–19 | 844 | 87 (10.3) | 1.00 (ref) | 1.00 (ref) d
| ||
20–22 | 1646 | 142 (8.6) |
0.76 (0.6–0.99)
| 0.78 (0.6–1.02) | |||
≥ 23 | 787 | 68 (8.6) | 0.80 (0.6–1.1) | 0.84 (0.6–1.2) | |||
Gallagher
2011
CHD [28] | <20 | 12,460 | 43 (0.35) | 0.93 (0.7–1.3) | D: age; SE: - BM: - HB: - R: - | Young age at first birth category not compared to lowest risk category. | |
20–24 | 72,570 | 239 (0.33) | 1 (ref) | ||||
25–29 | 124,044 | 101 (0.08) |
0.75 (0.6–0.95)
| ||||
≥ 30 | 45,041 | 37 (0.08) | 0.85 (0.6–1.2) | ||||
Gallagher
2011
Ischemic stroke [28] | <20 | 12,460 | 75 (0.60) | 1.23 (0.96–1.6) | |||
20–24 | 72,570 | 309 (0.43) | 1 (ref) | ||||
25–29 | 124,044 | 141 (0.11) | 0.88 (0.7–1.1) | ||||
≥ 30 | 45,041 | 50 (0.11) | 0.95 (0.7–1.3) | ||||
Gallagher
2011
Hemorrhagic
Stroke [28] | <20 | 12,460 | 178 (1.43) | 1.09 (0.9–1.3) | |||
20–24 | 72,570 | 850 (1.17) | 1 (ref) | ||||
25–29 | 124,044 | 423 (0.34) |
0.84 (0.7–0.9)
| ||||
≥ 30 | 45,041 | 132 (0.29) |
0.81 (0.7–0.97)
| ||||
Webb
2011 [10] | N/A | 1,021,417 b
| 46 (0.0045) | 1 (ref) | D: age, decade; SE: - BM: - HB: - R: - | Few events in <20 category. Nulliparous group includes young women who did not have children yet. | |
<20 | 159,716 | 16 (0.0100) | 1.25 (0.7–2.2) | ||||
≥ 20 | 732,462 | 71 (0.0097) | 0.93 (0.6–1.4) | ||||
<20 | 159,716 | 16 (0.0100) | 1.35 (0.8–2.3) | ||||
≥ 20 | 732,462 | 71 (0.0097) | 1 (ref) | ||||
Merritt
2015
(FTP)
(CVD) [32] | <21 | 39,201 | 304 (0.78) | 1.15 (0.99–1.3) | D: age, site; SE: education; BM: BMI; HB: physical activity, smoking (duration and intensity); R: menopausal status. | Extensive adjustment in final model and no unadjusted model for comparison. | |
21–23 | 71,322 | 497 (0.70) | 1.11 (0.98–1.3) | ||||
24–25 | 52,056 | 359 (0.69) | 1.05 (0.9–1.2) | ||||
26–30 | 75,927 | 557 (0.73) | 1 (ref) | ||||
≥ 31 | 27,027 | 237 (0.88) | 1.06 (0.9–1.2) | ||||
Merritt
2015
(FTP)
(CeVD) [32] | <21 | 39,201 | 97 (0.25) | 1.14 (0.9–1.5) | |||
21–23 | 71,322 | 150 (0.21) | 0.96 (0.8–1.2) | ||||
24–25 | 52,056 | 122 (0.23) | 1 (ref) | ||||
26–30 | 75,927 | 199 (0.26) | 0.97 (0.8–1.2) | ||||
≥ 31 | 27,027 | 79 (0.29) | 0.90 (0.7–1.2) | ||||
Merritt
2015
(FTP)
(CHD) [32] | <21 | 39,201 | 105 (0.27) | 1.14 (0.9–1.5) | |||
21–23 | 71,322 | 140 (0.20) | 0.96 (0.7–1.2) | ||||
24–25 | 52,056 | 106 (0.20) | 1 (ref) | ||||
26–30 | 75,927 | 166 (0.22) | 0.99 (0.8–1.3) | ||||
≥ 31 | 27,027 | 74 (0.27) | 1.10 (0.8–1.5) | ||||
Barclay 2016
Women [22] | 15–19 | 2998 | 1148 (38.3) |
1.18 (1.1–1.3)
|
1.25 (1.1–1.4)
| D: birth cohort [2]; SE: (sibling fixed effect model) age of person’s mother at time of their own birth, attained socioeconomic status, educational attainment, marital status; BM: - HB: - R: Completed parity | |
20–24 | 5232 | 1894 (36.2) |
1.11 (1.02–1.2)
|
1.23 (1.1–1.4)
| |||
25–29 | 3052 | 1022 (33.5) | 1 (ref) | 1 (ref) | |||
30–34 | 1010 | 322 (31.9) | 0.94 (0.8–1.1) | 0.89 (0.7–1.1) | |||
35+ | 343 | 120 (35.0) | 1.14 (0.9–1.4) | 1.09 (0.8–1.5) | |||
Barclay 2016
Men [22] | 15–19 | 1469 | 661 (45.0) |
1.45 (1.3–1.6)
|
1.37 (1.2–1.6)
| ||
20–24 | 9380 | 3595 (38.3) |
1.11 (1.1–1.2)
|
1.10 (1.03–1.2)
| |||
25–29 | 9533 | 3395 (35.6) | 1 (ref) | 1 (ref) | |||
30–34 | 4590 | 1623 (35.4) | 0.98 (0.9–1.04) | 1.03 (0.9–1.1) | |||
35+ | 2211 | 793 (35.9) | 0.98 (0.9–1.1) | 0.97 (0.9–1.1) | |||
Parikh 2016 [35] | Nullipara | 10,462 | Data not provided. Total cases: 4607 | 1.00 (0.9–1.1) | D: age at enrollment; SE: income, education, neighborhood SES variables; BM: history of high cholesterol requiring pills, hypertension, diabetes; HB: smoking; nr of still births, nr of miscarriages, breastfed for >1 month | Extensive adjustment in final model | |
<20 | 8780 |
1.65 (1.5–1.8)
|
1.27 (1.1–1.4)
| ||||
20–24 | 29,803 |
1.25 (1.2–1.4)
|
1.14 (1.1–1.2)
| ||||
≥ 25 | 23,937 | 1 (ref) | 1 (ref) |
Mortality
Coronary heart disease (CHD)
CHD overall
Cerebrovascular
Stroke overall
Ischemic stroke, hemorrhagic stroke and subarchnoid haemorrhage seperately
Events
Incident diagnosis of acute non-fatal myocardial infarction
Presence of heart disease
Mixed outcomes (mortality and events)
Highlighting important methodological differences and limitations
Operationalization of exposure categories
Choice of reference category
Sample size, age at follow-up and follow-up time
Confounding variables and proposed pathways
Discussion
Summary of the literature examining associations between AFB and CVD
Methodological limitations and their potential implications
Operationalization of exposure categories
Choice of reference category
Sample size, age of the cohort and follow-up time
Confounding variables and proposed pathways
Study strengths and limitations
Recommendations to improve the state of the literature
|
- Investigate the distributional tails of AFB, especially very early and very late age at first birth. - Justify AFB exposure categorization. How were categories selected (theory, biological evidence, statistical considerations, etc.)? - Reference category selection should reflect underlying hypotheses about exposed and unexposed groups. - Assure sufficient sample size and follow up time, taking into account cohort-age (fewer events should be expected for younger cohorts). For smaller studies, a priori power calculations are advisable. - Improved use of theory to direct covariate choice and modelling approaches. Where appropriate, future research would benefit from more sophisticated use of tools for investigating pathways (e.g. directed acyclic graphics, mediation modelling, structural equation modelling, etc.) - Ideally, prospective studies that start well before reproductive age could help answer questions about selection effects (for example, are observed associations between early AFB and CVD linked to some common earlier life event such as obesity in childhood or a “risk-taking” personality) |