The association of depressive symptoms and smoking during and after pregnancy: a longitudinal study

1. The association of depressive symptoms
and smoking during and after pregnancy:
a longitudinal study
Katrien De Wilde, PhDc
KAHO Sint-Lieven, Belgium

2. Title after revision
Smoking patterns, depression and
socio-demographic variables among
Flemish Women during Pregnancy
and the Postpartum period
2

3. Introduction
• Prevalence of smoking during pregnancy in Flanders:
12.3%
(Hoppenbrouwer et al. , 2011)
• Smoking during pregnancy
– Fetal and maternal morbidity
(Levitt et al., 2007, Lumley et al., 2009)
– Influence on feelings of dysphoria and
depression (Park et al., 2009; Scott et al., 2009)
• Depression during pregnancy
– Pre-term birth (Grote et al., 2010)
– Progression to postpartum depression (Bennett et al., 2004)
– Misinterpretation of symptoms
– Persistent smoking during pregnancy (Scott et al., 2009)
3

4. Aims of the study
1) Which patterns of smoking behavior can be observed
during and after pregnancy?
2) Which patterns of depression can be observed during
and after pregnancy?
3) Is there a relationship between smoking pattern and
feelings of depression during and after pregnancy,
independently from socio-demographic characteristics?
4

5. Method
• Observational, prospective, non-interventional study
• Data collection on 3 moments:
T0: < 16weeks pregnancy
T1: 32 – 34 weeks pregnancy
T2: > 6 weeks postpartum
• Questionnaire:
Smoking behavior (partner)
Socio-demographic variables: age, gravidity,
educational level, job status
Beck Depression Inventory (BDI) (Beck et al., 1979)
5

6. Results: smoking patterns
6
Smoking
pattern
Explanation Number of
respondents
Attrition
Smokers Persistent smokers 53 35
Non-smokers Never smoking OR quit >
1 year prior to T0
416 32
Recent ex-
smokers
Non-smokers quit ≤ 1
year prior to T0
30 15
Initial smokers Smokers who quit at T1
or T2
14 0
Initial non-
smokers
Relapsed at T1 or T2 10 0
Total 523 82

7. Results: patterns of depression
7
Time point Smokers
Non-
smokers
Recent
ex-smokers
Initial
smokers
Initial non-
smokers
T0
Mean BDI 11.17 6.09 6.95 10.77 7.71
Range 0 - 36 0 - 24 0 - 29 3 - 19 3 - 13
T1
Mean BDI 11.28 7.45 7.02 12.33 6.00
Range 0 - 42 0 - 31 0 - 33 4 - 19 3 - 10
T2
Mean BDI 9.61 4.22 5.29 7.69 5.56
Range 0 - 31 0 - 13 0 - 21 0 - 18 1 - 13

8. Results of univariable analysis
• Significant higher BDI-scores:
– Smokers and initial smokers
(F(4,486) = 12.06; p < 0.001)
– Low educated women
(F(1,496) = 40.39; p < 0.0001)
– Women without a job
(F(1,489) = 6.14; p = 0.0136)
– During pregnancy
(F(2, 643) = 40.15; p < 0.0001)
– Having a smoking partner
(F(1, 495) = 5.37; p < 0.021)
– Women younger than 29 years
(F(1, 495) = 7.11; p < 0.0079)
8

9. Results of multivariable analysis
9
Low educational level High educational level
Mean BDI score according to smoking pattern and
educational level at T0, T1 and T2 (F(8,619) = 2.29; p = 0.02)

10. Conclusions
• Recent ex-smokers reported less symptoms of
depression compared to smokers and initial smokers,
independent of their educational level, suggesting
that smoking cessation shortly before or in early
pregnancy does not aggravate depressive symptoms
during pregnancy and in postpartum.
• Mean BDI scores decreased in postpartum, except in
low educated smokers, where BDI scores remained
constantly above ten during pregnancy and
postpartum, suggesting that smoking could be a way
of coping with difficult life conditions.
10

11. Conclusions
• Ex-smokers and initial smokers had the highest
percentage of smoking partners (71.1 and 71.4%
respectively).
This means that women who recently attempted
quitting, are more at risk for relapse and that their
partner should be involved in smoking cessation
counseling.
11

12. Limitations
• Relatively high drop-out rate of 13.56%
• Partially due to our own strict exclusion criteria for
smokers (39.8%) and recent ex-smokers (33.3%)
12

13. Implications for practice
• Take enough time to explore not only obstetric
parameters, but also the lifestyle of the pregnant
woman during a first consultation.
• Identify those women who may need more
specialized care and offer smoking cessation
counseling tailored to their needs and possibilities.
• Consider measuring the level of depression in
pregnant women, e.g. by using the BDI.
• If possible, involve the partner.
13

14. Acknowledgements
• Supervisors:
– Prof. L. Maes, PhD, UGhent
– Prof. M. Temmerman, PhD, MD, UGhent
– Prof. H. Boudrez, PhD, UGhent
– L. Trommelmans, PhD, KAHO Sint-Lieven
– H. Laevens, PhD, KAHO Sint-Lieven
Further information:
katrien.dewilde@kahosl.be
14