Smoking cessation, weight change and Coronary Heart Disease among postmenopausal women with and without diabetes

To the Editor

Cigarette smoking is an important cause of cardiovascular disease, and smoking cessation reduces the risk.^1-2 However, weight gain after smoking cessation may increase the risk of diabetes and weaken the benefit of quitting.³ One study⁴ found an association between smoking cessation and a lower risk of cardiovascular events among participants without diabetes that was not modified by weight gain. However, this study had limited power for participants with diabetes and the more specific outcome of coronary heart disease (CHD). We used data from the Women’s Health Initiative (WHI)⁵ to assess the association between smoking cessation, weight gain, and subsequent CHD risk among postmenopausal women with and without diabetes.

Methods

In the WHI, 161,808 post-menopausal women ages 50-79 years were recruited from 40 sites between 1993 and 1998 and followed up every 6-12 months. Loss to follow up was 3% to 5%. Smoking status was defined by self-report at baseline and year 3. Never smokers and former smokers did not smoke at either time point, continuous smokers smoked at both time points, and new quitters smoked at baseline but not at year 3. Women without known CHD at year 3 were followed up until CHD diagnosis, date of death, loss to follow-up or September 30, 2010, whichever occurred first. CHD cases were adjudicated by physicians and defined as the first occurrence of clinical myocardial infarction, silent myocardial infarction or death due to CHD.

Diabetes was self-reported and defined as either having diabetes at baseline or year 3. Self-reported diabetes has been validated as an indicator of diagnosed diabetes.⁶ Cox proportional hazards regression models were used to estimate hazard ratios (HRs) for risk of CHD in relation to smoking status stratified by diabetes status and weight gain between baseline and year 3 (categorized as <5kg, 5−<10kg or ≥10kg), with and without additional adjustment for covariates. Statistical analyses were performed in SAS 9.3. The WHI study was approved by institutional review boards at all participating centers. All participants gave written informed consent.

Results

Of 104,391 women followed, 3,381 developed CHD over a mean (SD) of 8.8 (2.8) years. Baseline characteristics are in Table 1. The incidence rates of CHD per 1000 person-years were 3.3 in never smokers, 3.7 in former smokers, 7.6 in continuing smokers, and 5.3 in new quitters. Among 98,053 women without diabetes, new quitters had an adjusted HR for CHD of 0.74 (95% CI: 0.57-0.95) compared with continuing smokers; for former smokers, the adjusted HR was 0.39 (95% CI: 0.34-0.45). Among 6,338 women with diabetes, new quitters had a lower risk for CHD (HR=0.36, 95% CI: 0.17-0.78), as did former smokers (HR= 0.41, 95% CI: 0.29-0.59) compared with continuing smokers. These associations were unchanged after further adjustment for weight change (Table 2).

Table 1.

Characteristics of participants at baseline and year 3 (n=104,391) by smoking status between baseline and year 3 follow-up.

Variable	Never smokers	Former smokers	Continuing smokers	New quitters
Total number of women	54,801	42,912	4,787	1,891
Age at baseline, mean (SD), years	63.1 (7.2)	62.7 (6.9)	60.8 (6.7)	61.0 (6.6)
White, non-Hispanic-ethnicity, No. (%)	44,836 (81.8)	37,264 (86.8)	3,787 (79.1)	1,512 (80.8)
Body Mass Index at baseline, mean (SD), kg/m2	27.7 (5.7)	27.9 (5.8)	26.7 (5.5)	27.3 (5.7)
Diabetes at year 3 (yes), No. (%)	3,441 (6.3)	2,495 (5.8)	252 (5.3)	150 (7.9)
Weight gain between baseline and year 3, mean (SD), kg	0.3 (9.8)	0.4 (9.8)	−0.04(9.7)	3.0 (10.8)

Table 2.

Association between smoking cessation and CHD events (n=104,391)

		Overall			Weight gain < 5 kgd			Weight gain 5-<10 kgd			Weight gain ≥ 10 kgd
Exposure	CHD cases	Model 1 HR (95%CI )a	Model 2 HR (95%CI )b	Model 3 HR (95%CI)c	CHD Cases	Model 1 HR (95%CI) a	Model 2 HR (95%CI) b	CHD Cases	Model 1 HR (95%CI)a	Model 2 HR (95%CI)b	CHD Cases	Model 1 HR (95%CI)a	Model 2 HR (95%CI)b
No diabetes by year 3
Continuing smokers	260	1	1	1	210	1	1	19	1	1	6	1	1
New quitters	76	0.73 (0.56 0.94)	0.74 (0.57 0.95)	0.74 (0.57 0.96)	48	0.73 (0.54 1.00)	0.75 (0.55 1.03)	16	0.95 (0.49 1.84)	0.95 (0.48 1.88)	7	0.65 (0.22 1.94)	0.79 (0.24 2.62)
Former smokers	1207	0.39 (0.34 0.45)	0.39 (0.34 0.45)	0.40 (0.35 0.46)	999	0.39 (0.34 0.45)	0.39 (0.34 0.46)	104	0.54 (0.33 0.89)	0.47 (0.28 0.79)	24	0.32 (0.13 0.78)	0.35 (0.13 0.98)
Never Smokers	1346	0.33 (0.29 0.37)	0.32 (0.28 0.37)	0.33 (0.29 0.38)	1136	0.33 (0.28 0.38)	0.32 (0.28 0.38)	95	0.41 (0.25 0.68)	0.36 (0.21 0.61)	31	0.34 (0.14 0.83)	0.35 (0.13 0.97)
Diabetes at year 3
Continuing smokers	39	1	1	1	35	1	1	3	1	1	1	1	1
New quitters	8	0.37 (0.17 0.79)	0.36 (0.17 0.78)	0.36 (0.16 0.77)	5	0.34 (0.13 0.86)	0.33 (0.13 0.87)	3	0.66 (0.13 3.30)	0.26 (0.03 2.02)	0	-	-
Former smokers	203	0.43 (0.30 0.60)	0.41 (0.29 0.59)	0.41 (0.29 0.59)	150	0.35 (0.24 0.51)	0.34 (0.23 0.50)	20	0.55 (0.16 1.86)	0.26 (0.06 1.03)	8	0.63 (0.08 5.17)	-
Never Smokers	242	0.36 (0.26 0.50)	0.33 (0.23 0.47)	0.33 (0.23 0.47)	197	0.32 (0.22 0.46)	0.30 (0.21 0.45)	20	0.41 (0.12 1.40)	0.17 (0.04 0.73)	11	0.75 (0.09 5.89)	-

^aModel 1 adjusted for only age at enrollment (<55, 55-59, 60-64, 65-69, 70-74, and ≥75 years).

^bMultivariable model 2 adjusted for age at enrollment (<55, 55-59, 60-64, 65-69, 70-74, and ≥75 years); race/ethnicity (American Indian or Native Alaskan, Asian or Pacific Islander, black or African American, Hispanic/Latino, non-Hispanic white, and other); education (high school or less, some college/technical training, college or some post-college, and master’s degree or higher); body mass index (calculated as weight in kilograms divided by height in meters squared) (<18.5, 18.5-24.9, 25.0-29.9, 30.0-34.9, 35.0-39.9, and ≥40); waist circumference (in continuous cm); physical activity as (<5, 5-<10, 10-<20, 20-<30, and ≥30 metabolic equivalent tasks per week); alcohol intake (nondrinker, past drinker, <1 drink/mo, 1 drink/mo-<1 drink/wk, 1-<7 drinks/wk, and ≥7 drinks/wk); hypertension (yes, no); high cholesterol level requiring pills (yes, no); and participation in different WHI study subcohorts (observational study or clinical trials and different treatment assignments for all 4 clinical trials).

^cMultivariable model 3 adjusted for all factors in footnoteplus weight change from baseline to year 3 (weight stayed within ±2.5 kg, weight gain 2.5-<5 kg, weight gain ≥5 kg, weight loss 2.5-<5 kg, and weight loss ≥5 kg).

^dNumber of cases with weight gain <5 kg, 5-<10kg and >=5 kg doesn’t sum to total because of missing weight at baseline or Year 3 in 233 cases.

Among women who gained less than 5 kg, the association between smoking status and CHD risk was similar to the overall results in both women with and without diabetes (Table 2). Among women who gained 5−<10kg or ≥ 10 kg without diabetes, former smokers had a lower CHD risk than continuing smokers; in women with diabetes, neither new quitters nor former smokers had a lower CHD risk.

Discussion

In this study, smoking cessation was associated with a lower risk of CHD among postmenopausal women with and without diabetes. Weight gain following smoking cessation weakened this association, especially for women with diabetes who gained 5 kg or more, although power was limited in this subgroup due to the small number of cases. Other limitations of the study are that it included only postmenopausal women and did not account for further changes in smoking, weight or diabetes status after year 3.