Results
In the present study, 74 141 women with a median follow-up of 8.7 years
(IQR: 8.1–9.3; total follow-up: 634 298 person-years) were finally
included in the analysis. The baseline characteristics of participants
included or excluded are shown in Table S3. The combined RRS showed a
slight statistical improvement in predicting CVD compared to isolated
items of RRS (Table S4–S5).
The distribution of RRS was right-skewed (Figure S1). The baseline
characteristics of participants by RRS groups are shown in Table 1. The
participants were divided into the low-risk group (n = 11 735,
15.8%), low-intermediate group (n = 38 898, 52.5%),
high-intermediate group (n = 19 258, 26.0%) and high-risk group
(n = 4250, 5.7%). As shown in Figure S1, the RRS showed a
decreasing trend with age. Participants in the healthy lifestyle group
(RRS: 4–5) had a lower mean RRS at age ≥ 45 years. The RRS was higher
in participants who were normal weight or overweight; had a higher level
of education; had lower alcohol intake equivalent and had normal
systolic blood pressure (120–140 mmHg) (Figure S2).
The incidence of CVD, IHD and stroke during follow-up were 3.8%
(n = 2792), 3.0% (n = 2214) and 0.9% (n = 657),
respectively. With the low-risk group (RRS: 0–1) as the reference, the
high-risk group (RRS: 6–13) had higher risks of CVD (adjusted hazard
ratio [aHR]: 2.07, 95% CI: 1.74 to 2.45), IHD (2.13: 1.76 to 2.57)
and stroke (1.81: 1.24 to 2.65). The adjusted PAR% were 16 (95% CI: 8
to 24) for CVD, 15 (95% CI: 6 to 24) for IHD and 18 (95% CI: 1 to 33)
for stroke (Table 2). The differences in cumulative rates between the
RRS groups increased over time. Nevertheless, the cumulative curve of
stroke risk could not clearly distinguish between different RRS groups
before 7.5 years of follow-up (Figure 1). The contribution of RRS to the
risk of CVD decreased in participants who maintained a healthy lifestyle
(Figure 2A). The combined PAR% in different RRS groups concerning HLS
was significant in CVD, with 26.3% (95% CI: 11.3% to 38.7%) in the
group with score 0–2, 26.3% (95% CI: 13.3% to 37.4%) in the group
with score 3–5 and 50.0% (95% CI: 22.8% to 67.7%) in the group with
score ≥ 6. Similar results were found for IHD, with 30.9% (95% CI:
14.9% to 43.9%) in the group with score 0–2, 31.6% (95% CI: 18.0%
to 43.0%) in the group with score 3–5 and 53.5% (95% CI: 24.9% to
71.3%) in the group with score ≥ 6. We also observed a similar tendency
for HLS. A healthy lifestyle can reduce the risks of CVD and IHD
attributed to RRS by about 20% compared to an unhealthy lifestyle
(Figure 2B).
The attenuated risk effects of categorical or continuous HLS in
different groups of RRS compared with the unhealthy lifestyle group
(HLS: 0–1) are shown in Table 3. In the high-risk group (RRS: 6–13),
the healthy lifestyle group with HLS 4–5 showed reductions of about
57% (HR: 0.43, 95% CI: 0.21 to 0.86) for CVD and 68% (HR: 0.32, 95%
CI: 0.13 to 0.76) for IHD, and each 1 point increase in the HLS
decreased the CVD risk by 32% (HR: 0.68, 95% CI: 0.52 to 0.89) and the
IHD risk by 35% (HR: 0.65, 95% CI: 0.49 to 0.88). There was no
significant association between a healthy lifestyle and the occurrence
of stroke.
In a joint analysis of RRS and HLS, each increase in HLS group reduced
the CVD and IHD risks compared with the low-risk RRS group (0–1) and
unhealthy lifestyle group (0–1). Even in the high-risk subgroup (RRS:
6–13), the benefits of a healthy lifestyle outweighed the negative
effects of poor reproductive status, with HR of 0.24 (95% CI: 0.16 to
0.38) for CVD and HR of 0.18 (95% CI: 0.11 to 0.32) for IHD (Figure 3).
The RRS showed more pronounced associations with CVD, IHD and stroke
among older participants (age ≥ 60 years) (aHR: 1.09, 95% CI: 1.06 to
1.12 for CVD; aHR: 1.09, 95% CI: 1.05 to 1.13 for IHD; aHR: 1.08, 95%
CI: 1.02 to 1.15 for stroke). Analyses stratified by age group and
systolic blood pressure showed significant associations of the RRS with
CVD and IHD (P interaction = 0.005 in age group
for CVD, P interaction = 0.002 in age group andP interaction = 0.03 in systolic blood pressure
group for IHD) (Table S6).
Additive interaction effects of RRS and HLS on the risks of incident
CVD, IHD and stroke were also detected. The relative excess risks of
interaction were 0.05 (95% CI: 0.01 to 0.09, P < 0.05)
for CVD and 0.06 (95% CI: 0.01 to 0.011, P < 0.05) for
IHD (Table 4). Attributable proportions due to interaction of about 14%
(95% CI: 7% to 22%) for CVD and 15% (95% CI: 9% to 23%) for IHD
were observed, indicating that about 15% of cases were due to the
interaction effect between RRS and HLS. However, no additive interaction
effect was found between RRS and HLS in stroke.