Smoking and Depression Risk Reduction in a Primary Care Setting

Abstract

Introduction:

Smoking is associated with many diseases and is a target for primary preventive efforts in numerous morbidities. Studies show that smoking and depression may be associated. Never-smokers are at significantly lower risk than current and former smokers. Despite this observation, the effect of smoking on depression risk reduction has not been adequately explored. The purpose of this study was to explore the effect of smoking on depression risk reduction in adult patients seen in a primary care clinic at an academic medical center 6 months after they were identified as being at risk for depression. Findings may influence the direction and intensity of our smoking cessation endeavors in patients at risk of depression who smoke.

Methods:

We conducted an analytic cross-sectional study using electronic medical records of patients 18 years and older seen a primary care setting between January 1, 2019 and December 31, 2020. All participants included had an initial depression risk score (assessed by the 9-item Patient Health Questionnaire (PHQ-9)) of 5 or higher, information on smoking status and a PHQ-9 score at 6 months. We determined the percentage of patients with PHQ-9 score decrease of 5 or more at 6 months and used logistic regression to determine the association depression risk reduction (of 5 units or more) at 6 months and smoking, adjusting for demographic, clinical, and behavioral characteristics.

Results:

Number of patients included were 120, mean age was 55 (16), years, 88 (74%) were female, 68 (57%) were African American, and 31 (26%) were Caucasian. Fifty (44%) had a history of smoking and 31 (25.8%) had improvement (ie, a decrease of 5 units or more) in their PHQ-9 score at 6 months. Smoking was associated with decreased odds of improvement in depressive symptoms (Odds ratio = 0.32, 95% Confidence interval: 0.12-0.87).

Conclusion:

Risk of depression was more likely to persist in smokers than non-smokers at 6 months. Addressing smoking behavior in those with risk of depression may be beneficial.

Introduction

Smoking is a risk factor for many diseases including depression. Depression is an important cause of morbidity around the world. The World Health Organization (WHO) estimates that it affects nearly 5% of all adults. ¹ In the United States, prevalence of major depressive episode among adults 18 years and older was 8.3%, estimated to be 21 million adults in 2021. ² Smoking cessation reduces the risk of many poor health outcomes. ³ Persons diagnosed with current mental health disorders have a higher prevalence of smoking than those with no mental health disorder (39% vs 16%), and it is estimated that while adults with mental illness represent 25% of the total population, they consume over 40% of all cigarettes smoked.^4,5 Smokers are not only at an elevated risk of depression but also of suicidal behavior independent of prior depression, making it very important to focus on improvement in depression among this vulnerable group of adults. ⁶ A study comparing current smokers, former smokers and never smokers showed that risk of depression was significantly lower with never-smokers than current and former smokers.^{7 -9} Reasons for the association between smoking and depression may include their shared environmental risk factors (eg, socio-economic factors including poverty, unemployment, and home instability) as well genetic factors. ¹⁰ Prevalence of both smoking and depression are higher in disadvantaged populations such as people with low socioeconomic status.^10,11 Although association between smoking and depression has been established in many studies, smokers differ in their perceptions on quitting smoking as an intervention for depression This stems from a notion among users that smoking cessation might worsen their depression.^12,13

While associations between depression and smoking are well-documented, the effect of smoking on depression risk reduction has not been adequately explored. We describe the effect of smoking on depression risk in adult patients seen in a primary care clinic at an academic medical center 6 months after they were identified as being at risk for depression. The findings could influence the direction and intensity of our smoking cessation endeavors in patients at risk of depression who smoke.

Methods

We conducted a cross-sectional analytic study involving review of longitudinally collected patient medical records. ¹⁴ Records of patients 18 years and older, seen at an academic family medicine clinic between January 1, 2019 and December 31, 2020 were considered. Records of patients were included if they had a score of at least 5 on the 9-item Patient Health Questionnaire (PHQ-9) ¹⁵ at baseline and a PHQ-9 score at 6 months. Patients with no information on smoking status at baseline were excluded from the analysis. Smoking status was defined as whether or not patients were smoked tobacco at baseline. This was extracted from the medical records and was indicated as “Yes,” if currently smoking or “No,” if otherwise. The 9-item patient health questionnaire (PHQ-9) is a common tool used by primary care providers to assess depression risk in their patients. Each question assesses how often in the past 2 weeks the patient had been bothered by certain feelings or thoughts. Each statement is scored on a scale of 0 to 4, 0 being “Not at all,” 1, “Several days,” 2, “More than half the days” and 3, “Nearly every day,” with a total score of 0 to 27, where higher scores denote more serious risk of depression. ¹⁵ Patients are routinely screened for depression at our clinic and those identified as being at risk of depression are either managed by the primary care provider or referred to mental health providers as appropriate. We collected PHQ-9 scores at baseline and 6 months. Where 6-month PHQ-9 scores were not available, we used PHQ-9 scores within 1 month and closest to 6 months in place of the 6-month scores (ie, 5-7 months), if available. We defined significant improvement in depression at 6 months as a decrease in PHQ-9 scores of 5 points or more from baseline.^16,17 We also collected data on demography (age, sex, and race/ethnicity), presence of clinical covariates (schizophrenia, diabetes, and hypertension), and behavioral characteristics (alcohol use). Data on these covariates were collected because of their known influence on depression.^{18 -22}

Data Analysis

We determined the percentage of patients who had significant improvement over their baseline PHQ-9 score (by our definition of a decrease in PHQ-9 score of 5 or more) at 6 months of follow-up.

We described the demographic, clinical, and behavioral characteristics of included participants using frequencies and percentages. Age was expressed as a mean and standard deviation. We compared the covariates between those who improved in their PHQ-9 scores and those who did not at 6 months follow-up using chi-square for categorical variables and t-test for continuous variables. Logistic regression analysis was then performed to determine independent associations between improvement in PHQ-9 scores (ie, decrease in depression risk) at 6 months of follow-up and smoking status at baseline, adjusting for the demographic, clinical, and behavioral covariates. We expressed our results in odds ratios (OR) and 95% confidence intervals (95% CI) with a level of significance of .05.

The study was determined to qualify for exempt status by our Institutional Committee for the Protection of Human Subjects as it used existing data.

Results

A total of 504 unique patients were identified at baseline. One hundred and twenty patients who had information on smoking status at baseline in addition to demographic and clinical covariates were selected for further analysis. Mean age was 55 ± 16 years with a range of 18 to 89 years. Eighty-nine (74%) were female, 68 (57%) were African American, 31 (26%) were Caucasian, and 50 (44%) had a history of smoking. Table 1 shows the details of the patient characteristics. Thirty-one (25.8%) of the participants saw improvements (ie, a decrease of 5 units or more from baseline) in their PHQ-9 scores at 6 months. Table 2 compares the characteristics of patients who saw improvements in their PHQ-9 scores and those who did not at 6 months. Significantly fewer current smokers improved in their PHQ-9 scores compared to non-smokers (P = .05).

Table 1.

Characteristics of Study Population at Baseline.

Covariate	Number, N = 120 (%)
Sex
Female	89 (74.2)
Male	31 (25.8)
Race
Caucasian	31 (25.8)
African American	68 (56.7)
Hispanic or Latino	8 (6.7)
Other	13 (10.8)
Schizophrenia
No	116 (96.7)
Yes	5 (3.3)
Alcohol use
No	75 (62.5)
Yes	45 (37.5)
Smoking status
No	67 (55.8)
Yes	53 (44.2)
Diabetes
No	76 (63.3)
Yes	44 (36.7)
Hypertension
No	36 (30.0)
Yes	84 (70.0)

Average age: 55.0; Standard deviation (SD): 16.31; Range: 18-89 years.

Table 2.

Comparison of Characteristics by Improvement in PHQ-9 Scores Status at 6 Months.

Characteristic	Improved	Did not improve	P-value
	N = 31 (%)	N = 89 (%)
Mean age (±SD), years	55.2 (15.8)	55.0 (16.6)	.96
Sex
Male	8 (25.8)	23 (25.8)	1.00
Female	23 (74.2)	66 (74.2)	1.00
Race
Caucasian	6 (19.3)	25 (28.1)	.34
African American	20 (64.5)	48 (53.9)	.31
Hispanic	2 (6.4)	6 (6.7)	.96
Other	3 (9.7)	10 (11.2)	.81
Schizophrenia	2 (6.4)	8 (9.0)	.66
Smoking	9 (29.0)	44 (49.4)	.05*
Alcohol use	12 (38.7)	33 (37.1)	.87
Diabetes	12 (38.7)	32 (36.0)	.78
Hypertension	22 (71.0)	62 (69.7)	.89

Abbreviation: SD: standard deviation.

^*Statistically significant.

In a multivariable logistic regression analysis adjusting for all the covariates, smoking remained the only covariate that showed significant association with reduction in depressive symptoms at 6 months of follow-up, with a 68% decreased odds of improvement in depression symptoms after 6 months follow-up in smokers compared with non-smokers (adjusted OR = 0.32, 95% CI: 0.12-0.87). Table 3 shows the associations between the covariates and improvement in depression risk in the adjusted logistic regression model.

Table 3.

Factors Associated With Improvement in Depression Symptoms.

Covariate	Adjusted odds ratio	95% Confidence interval
Male (compared to female)	1.25	0.45-3.51
Black (compared to Caucasians)	2.16	0.71-6.59
Hispanic	1.33	0.186-9.58
Other	1.29	0.24-6.85
Age	1.00	0.97-1.03
Schizophrenia	1.21	0.10-15.24
Alcohol use	1.45	0.55-3.80
Smoking	0.32	0.12-0.87*
Diabetes mellitus	1.07	0.42-2.73
Hypertension	1.06	0.34-3.24

^*Statistically significant.

Discussion

We reviewed records of adult patients who had initially been identified as at risk for depression due to their PHQ-9 score of 5 or above and followed up for 6 months in a primary care setting. We found that 26% of the included patients showed an improvement with a decrease of PHQ-9 score of 5 or more at 6 months. Smoking was a significant factor that impeded risk reduction of depression in our study. In other words, the risk of depression was more likely to persist in smokers than non-smokers at 6 months.

Our finding may indirectly be in agreement with studies that have suggested beneficial effects of smoking cessation in improving depression.^{23 -25} A systematic review and meta-analysis by Taylor et al ²⁴ found that quitting smoking may have a moderate improvement in not only depression symptoms but also in anxiety and mixed anxiety and depression. Improvement in mood and quality of life with reduction of depression, anxiety, and stress have also been reported after smoking cessation in adults. ²³

While many studies have found associations between smoking and depression as well as other mental health conditions, the novelty of our study is that we looked at the association between smoking and its effect on depression risk reduction. The high frequency of co-occurrence of smoking and depression has led to speculations as to the reasons for the association. In some circles, smoking has been thought of as self-medication for stress reduction in people suffering from mental health illnesses, including major depression. It is believed that smokers addicted to nicotine may experience some withdrawal effect and smoking relieves them of the withdrawal effects. Another postulated mechanism for the association is decrease in serotonin associated with chronic nicotine exposure.^6,26

Our study has shown, that depression risk may be more persistent in smokers than non-smokers.

Although some studies found associations between chronic diseases, for example, hypertension and diabetes mellitus, and risk of or worsening depression,^27,28 our data did not show any significant associations between presence of chronic diseases and worsening depression risk. Our relatively small sample size might be partly accountable for that. Regardless, a negative smoking status on follow up remained an independent factor associated with an improved depression score.

Limitations

Our designation of smokers was based on smoking status at baseline. We had no data on how many were still smoking at 6 months as well as how much and for how long they had been smoking. We had no information on what kind of intervention patients identified received for depression or if any smoking cessation intervention was prescribed. We did not distinguish between commercial tobacco products such as cigarettes, e-cigarettes, and pipes. We had no data on factors that may contribute to the risk of depression such as social isolation, joblessness or difficult life situations.

Conclusion

Our study showed that smoking is associated with persistence of depression risk indicated by screening using PHQ-9 in primary care settings. Addressing smoking in patients with risk of depression would be beneficial, and primary care physicians can play a critical role in this intervention. Currently PHQ-9 is a United States Preventative Services Task Force (USPSTF) recommended screening tool to assess depression risk in primary care clinics. Identifying smokers among those deemed at risk of depression using a routine screening by primary care physicians may provide an early start of the intervention, which otherwise can become a complicated and costly health problem.