Abstract
Background:
Several studies have noted an association between tobacco smoke and depression. Cadmium and lead are neurotoxicant components of tobacco smoke. The objective of the present study is to investigate the potential association between BCd and BPb with current depressive symptoms in the US adult population.
Methods:
We conducted cross-sectional analyses of adult participants (≥20 years) from the National Health and Nutrition Examination Survey (NHANES) 2011–2012 (n=3905). Multivariate logistic regressions were used to analyze the association between blood cadmium (BCd) and blood lead (BPb) with depressive symptoms; analyses were also stratified on sex and age groups (20–47 years old, and, 48 and older) Presence or absence of depressive symptoms was determined using the Patient Health Questionnaire (PHQ-9) module.
Results:
Individuals in the highest quartile of BCd had higher odds of having depressive symptoms (OR=1.68; 95% CL: 1.12, 2.51). This association was found only in male participants, and more specifically, in younger adult male participants (20–47 years). We found that BPb, cigarette smoking, and obesity were associated with depressive symptoms in younger female adults.
Conclusion:
In this study, we report associations between BCd and BPb with current depressive symptoms that were modified by age and sex. Reverse causation cannot be ruled out as a possible explanation since depression may lead to behavioral changes that increase exposure to cadmium and lead (i.e. tobacco smoke). The continued efforts at reducing cadmium through tobacco smoking cessation programs may decrease the prevalence of current depressive symptoms.
INTRODUCTION
Depression is a common mental disorder with an estimated prevalence of 9.1% among U.S. adults (18 years of age and older) (http://www.cdc.gov/features/dsdepression/index.html). Moreover, depression has become an important cause of disability worldwide, which is also reflected in decreased work productivity.1 Genetic and environmental factors contribute to the risk of depression.2 Additionally, it has been suggested that sex differences play a role in the development of depression, with women having higher rates.3
Cigarette smoking is strongly associated with depression, and this association may be bidirectional: smoking increases the risk of depression,4,5 and on the other hand, depression increases the use of cigarette smoking as a self-medicating behavior.6,7 Although smoking is more common in male,8 depression is more commonly associated with smoking in women than men.9,10
More than 8,400 chemical constituents are present in tobacco and tobacco smoke.11 Among them are the heavy metals cadmium (Cd) and lead (Pb). Cadmium is a widespread industrial and environmental pollutant, with inhalation being the primary route of exposure.12 Moreover, Cd from cigarette smoking is a significant source of environmental exposure.12 Cadmium is also present in low amounts in food, but dietary intake and gastrointestinal absorption are minimal. 12 Cadmium toxicity affects several organs including kidney, lung, liver, and brain;12 an association between blood cadmium (BCd) and depressive symptoms was recently reported in young adults.13 Lead is a neurotoxicant,14 and the association between blood lead (BPb) and depression has shown inconsistent results with one study reporting a positive association15 while several others found no statistically significant association.13,16
The objective of the present study is to investigate the potential association between BCd and BPb with current depressive symptoms in the National Health and Nutrition Examination Survey (NHANES) 2011–2012. Additionally, we conducted stratified analyses to identify differences in risk factors associated with current depression symptoms by sex and age.
METHODS
Study population
NHANES is a cross-sectional, nationally representative survey of the non-institutionalized civilian population of the United States conducted annually by the National Center for Health Statistics (NCHS), CDC.17 For our study, we used the publicly available files for NHANES 2011–2012. The survey employs a multistage stratified probability sample based on selected counties, blocks, households, and persons within households.
NCHS-trained professionals conducted interviews and extensive physical examinations in participants’ homes and collected blood and urine samples at mobile exam centers. All procedures were approved by the NCHS Research Ethics Review Board (Continuation of Protocl #2011–17 http://www.cdc.gov/nchs/nhanes/irba98.htm), and all participants provided written informed consent. For our analyses, we included adult participants (≥20 years of age) who answered the Patient Health Questionnaire (PHQ)-9 module (n=3,905). Pregnant women and women breastfeeding were excluded, however, inclusion of these persons did not change the statistical significance of our analyses (data not shown). Additionally, participants with missing co-variables included in the multivariable-adjusted models were excluded for a final sample size of 3,903 participants.
Outcome Measure
The outcome was the presence or absence of depressive symptoms as determined by a participant’s score on the PHQ-9, a self-administered version of the depression module of the Primary Care Evaluation of Mental Disorders Questionnaire (PRIME-MD). 18 PHQ-9 contains nine questions about the frequency of symptoms of depression over the past 2 weeks that are used as a depression screener in NHANES 2011–2012. These are based on the nine signs and symptoms for depression listed in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). Responses to these nine questions were on a four-point Likert scale of 0 to three, indicating that the participant experienced the symptom “not at all,” “on several days,” “on more than half the days,” or “nearly every day” during the past two weeks for a total score ranging from 0 to 27. A prior validation study found that a score ten or higher achieved 88% sensitivity and 88% specificity for major depression.18 Therefore, a participant who scored ten or more was defined as having depressive symptoms.
Exposure measures
Whole blood cadmium (BCd) and blood lead (BPb) concentrations were measured using inductively-coupled plasma mass spectrometry by CDC’s National Center for Environmental Health (NCEH), Division of Laboratory Sciences (DLS). Detailed methodology and QA/QC instructions are discussed in the NHANES Laboratory Procedures Manual (http://www.cdc.gov/nchs/data/nhanes/nhanes_07_08/PbCd_E_met_lead_cadmium.pdf; http://www.cdc.gov/NCHS/data/nhanes/nhanes_09_10/PbCd_F_met.pdf)
BCd and BPb were categorized as weighted quartile based on the distribution of the metals levels among the study population. Participants who had BCd or BPb values below the limit of detection (LOD 2011–2012: 0.16 μg/l for BCd and 0.25 μg/dL for BPb) were assigned the limit of detection divided by the square root of two, as recommended by NHANES. The percentage of participants in the study with measurement below LOD were 14.17% for BCd and 0.66% for BPb.
Covariates
Models were adjusted for a priori factors based on previous literature demonstrating an association with depression.19–21 These include: age (categorized in weighted quartiles), sex, race/ethnicity, education, poverty income ratio, obesity, alcohol consumption, cigarette smoking, and serum cotinine as a biomarker of tobacco smoke exposure. We obtained information about age (years), sex, race/ethnicity, and education from the household interview. Race/ethnicity was categorized as non-Hispanic white, non-Hispanic black, Hispanic, and other. Poverty income ratio (PIR) is a measure of socioeconomic status and represents the calculated ratio of household income to the poverty threshold after accounting for inflation and family size. Body Mass Index (BMI) was obtained from the physical examination and was calculated by dividing measured weight in kilograms by measured height in meters squared.. The adult population was classified as normal/underweight, overweight, and obese with BMI measures of <25, 25–29.9, and ≥30 kg/m2, respectively. Alcohol consumption (amount consumed per week) and smoking information were obtained from the associated questionnaire. Smoking status was defined as non-smoker (smoke <100 cigarettes ever), former smoker (not currently smoking, but has smoked ≥100 cigarettes ever), and current smoker. Serum cotinine was categorized as weighted tertiles; because more than a third of the participants had serum cotinine levels below the limit of detection (LOD: 0.015 ng/mL), the lowest, referent tertile contained those participants with levels below the LOD. Furthermore, adding to our analyses chronic conditions such as self-reported diabetes, hypertension, self-reported cardiovascular disease (defined as an answer of yes to any of coronary artery disease, angina pectoris, heart attack, stroke, or congestive heart failure on the medical questionnaire) did not change the statistically significant association that we reported (data not shown).
Statistical Methods
The Mobile Examination Center (MEC) exam sample weights were used for analyses to account for the complex sampling design and non-response of NHANES (http://www.cdc.gov/nchs/data/nhanes/nhanes_09_10/mecinterviewers.pdf). We used logistic regression to calculate adjusted odds to have depressive symptoms; odds ratios and 95% confidence intervals are reported in the tables. Furthermore, we stratified our analyses on sex and age group (20–47 years of age and > 47 years of age); age-group stratification was based on the weighted median of the study population. SAS 9.3 (SAS Institute, Cary, NC) was used for all statistical analyses and SAS-Callable SUDAAN 10 (Research Triangle Institute, Research Triangle Park, NC) was used to account for the NHANES complex sample design. P-values from Satterthwaite statistics were presented at the significance level ≤ 0.05.
RESULTS
Table 1 presents the characteristics of the study population. The geometric mean age of all participants was approximately 44 years old, and women (average age of 45 years) were slightly older than men (average age of 43 years). Roughly seventy percent of the participants are either overweight or obese; the prevalence of obesity was higher in women (about 38%) than in men (34%). The geometric mean BCd levels for all, male, and female adult participants were 0.33 μg/L, 0.30 μg/L, and 0.37 μg/L, respectively; the geometric mean BPb levels for all, male, and female adult participants were 1.09 μg/dL, 1.29 μg/dL, and 0.93 μg/dL, respectively. Nearly 10% of all adult participants were categorized as having depressive symptoms, with a higher prevalence in women (11.5%) than in men (7.3%) (Table 1).
Table 1:
Sample size and weighted characteristics of adult participants in NHANES 2011–2012
| ALL | MEN | WOMEN | |
|---|---|---|---|
| n | 3905 | 1998 | 1907 |
| Men | 50.08 (0.90) | ||
| Women | 49.92 (0.90) | ||
| Blood Cadmium (μg/L), GM (SE) | 0.33 (0.01) | 0.30 (0.01) | 0.37 (0.01) |
| Blood Lead (μg/dL), GM (SE) | 1.09 (0.03) | 1.29 (0.04) | 0.93 (0.03) |
| Age (Years), GM (SE) | 44.33 (0.98) | 43.41 (0.96) | 45.27 (1.05) |
| BMI (kg/m2), GM (SE) | 28.16 (0.22) | 28.05 (0.23) | 28.28 (0.23) |
| Serum Cotinine (ng/mL), GM (SE) | 0.23 (0.03) | 0.40 (0.07) | 0.14 (0.02) |
| Depression | |||
| Yes, % (SE) | 9.38 (1.02) | 7.30 (0.99) | 11.46 (1.24) |
| No, % (SE) | 90.62 (1.02) | 92.70 (0.99) | 88.54 (1.24) |
| Body Weight Status | |||
| Underweight/Normal weight, % (SE) | 30.29 (1.76) | 28.10 (1.62) | 32.49 (2.05) |
| Overweight, % (SE) | 33.60 (1.44) | 37.47 (1.58) | 29.73 (2.01) |
| Obese, % (SE) | 36.11 (1.53) | 34.43 (1.51) | 37.79 (2.04) |
| Poverty Income Ratio | |||
| PIR (below or equal poverty line), % (SE) | 16.25 (1.63) | 15.61 (1.72) | 16.90 (1.66) |
| PIR Above poverty line, % (SE) | 83.75 (1.63) | 84.39 (1.72) | 83.10 (1.66) |
| Smoking Status | |||
| Current Smoker, % (SE) | 19.78 (1.11) | 23.17 (1.66) | 16.39 (1.37) |
| Former Smoker, % (SE) | 25.05 (1.48) | 27.22 (1.76) | 22.87 (2.13) |
| Never Smoked, % (SE) | 55.17 (1.50) | 49.61 (1.99) | 60.74 (1.92) |
| Alcohol Consumption | |||
| No Alcohol, % (SE) | 18.64 (0.98) | 10.11 (0.91) | 27.19 (1.64) |
| 1–4 drinks per week, % (SE) | 73.16 (1.26) | 76.61 (1.90) | 69.71 (1.48) |
| >4 drinks per week, % (SE) | 8.20 (1.05) | 13.29 (1.63) | 3.10 (0.57) |
| Education Level | |||
| Less than High School, % (SE) | 14.53 (1.79) | 15.52 (1.86) | 13.53 (1.84) |
| Completed High School, % (SE) | 20.26 (1.58) | 21.90 (1.99) | 18.63 (1.52) |
| More than High School, % (SE) | 65.21 (2.76) | 62.58 (3.09) | 67.84 (2.66) |
| Race/ethnicity | |||
| Non-Hispanic White, % (SE) | 69.64 (3.74) | 69.91 (3.78) | 69.37 (3.81) |
| Non-Hispanic Black, % (SE) | 10.26 (2.09) | 8.93 (1.93) | 11.59 (2.28) |
| Hispanic, % (SE) | 13.12 (2.45) | 13.89 (2.63) | 12.34 (2.33) |
| Other, % (SE) | 6.98 (1.01) | 7.27 (1.05) | 6.70 (1.05) |
Multivariate logistic regression analyses found that among all adult participants, those in the highest quartile of BCd were statistically significantly associated with higher odds to have depressive symptoms (OR=1.68, 95% CL: 1.12, 2.51) compared to those in the lowest referent quartile (Table 2). After stratification by sex, this statistically significant association was found in adult male participants (OR=2.59, 95% CL: 1.11, 6.00), but not in females (OR=1.58, 95% CL: 0.77, 3.24) (Table 2). Further stratification by age groups found that younger male adults (ages 20–47 years) in the highest quartile of BCd were statistically significantly associated with higher odds for having depressive symptoms (OR=3.16, 95% CL: 1.26, 7.91) compared to the lowest, referent quartile (Table 3); however, in older adult male participants (≥48 years) BCd was not significantly associated with higher odds for having depressive symptoms (Table 4). Interestingly, older adult female participants in the highest BCd quartile were statistically significantly associated with higher odds for having depressive symptoms (OR=3.30, 95% CL: 1.02, 10.69) compared to the referent quartile (Table 4).
Table 2.
Multivariate logistic regressionaOR (95% CL) of having depression in NHANES 2011–2012 adult participants.
| All | Men | Women | |
|---|---|---|---|
| n | 3903 | 1997 | 1906 |
| Depressed yes, n | 429 | 170 | 259 |
| Depressed no, n | 3474 | 1827 | 1647 |
| Sex | |||
| Men | 1.00 | ||
| Women | 1.78 (1.33, 2.37)*** | ||
| Race/Ethnicity | |||
| Non-Hispanic White | 1.00 | 1.00 | 1.00 |
| Non-Hispanic Black | 0.98 (0.64, 1.49) | 0.66 (0.40, 1.09) | 1.16 (0.59, 2.28) |
| Hispanic | 1.10 (0.66, 1.83) | 0.98 (0.56, 1.72) | 1.31 (0.68, 2.52) |
| Other | 1.05 (0.70, 1.57) | 0.71 (0.36, 1.38) | 1.54 (0.84, 2.82) |
| Age quartile | |||
| Age Q1 (20–32 years) | 1.00 | 1.00 | 1.00 |
| Age Q2 (33–47 years) | 1.19 (0.74, 1.90) | 0.85 (0.37, 1.96) | 1.45 (0.77, 2.70) |
| Age Q3 (48–60 years) | 1.55 (0.98, 2.46) | 1.85 (0.94, 3.64) | 1.41 (0.73, 2.70) |
| Age Q4 (> 60 years) | 0.78 (0.42, 1.43) | 0.57 (0.15, 2.12) | 0.91 (0.43, 1.92) |
| Education level | |||
| Less than High School | 1.00 | 1.00 | 1.00 |
| Completed High School | 0.87 (0.52, 1.46) | 0.97 (0.47, 2.00) | 0.79 (0.47, 1.30) |
| More than High School | 0.68 (0.37, 1.27) | 0.64 (0.31, 1.32) | 0.69 (0.37, 1.30) |
| Poverty Income Ratio | |||
| PIR ≤ 1 | 1.00 | 1.00 | 1.00 |
| PIR >1 | 0.41 (0.27, 0.63)*** | 0.40 (0.20, 0.82)* | 0.46 (0.30, 0.73)** |
| Body Weight Status | |||
| Underweight/Normal | 1.00 | 1.00 | 1.00 |
| Overweight | 1.11 (0.61, 2.01) | 0.97 (0.42, 2.21) | 1.16 (0.66, 2.05) |
| Obese | 1.73 (1.06, 2.82)* | 1.01 (0.47, 2.18) | 2.48 (1.39, 4.45)** |
| Alcohol consumption | |||
| No Alcohol | 1.00 | 1.00 | 1.00 |
| 1–4 drinks per week | 1.02 (0.71, 1.46) | 0.68 (0.37, 1.27) | 1.10 (0.70, 1.75) |
| >4 drinks per week | 1.06 (0.54, 2.06) | 0.52 (0.22, 1.20) | 2.07 (0.70, 6.15) |
| Smoking status | |||
| Current Smoker | 1.79 (1.19, 2.69)** | 0.85 (0.43, 1.67) | 3.50 (1.67, 7.31)** |
| Former Smoker | 1.46 (0.94, 2.28) | 1.28 (0.68, 2.44) | 1.76 (0.97, 3.20) |
| Never Smoked | 1.00 | 1.00 | 1.00 |
| Serum Cotinine Tertile | |||
| Serum Cotinine T1 | 1.00 | 1.00 | 1.00 |
| Serum Cotinine T2 | 0.76 (0.48, 1.22) | 0.52 (0.25, 1.08) | 0.93 (0.59, 1.45) |
| Serum Cotinine T3 | 1.01 (0.71, 1.45) | 1.23 (0.55, 2.74) | 0.79 (0.43, 1.45) |
| Blood Lead Quartile | |||
| BPb Q1 | 1.00 | 1.00 | 1.00 |
| BPb Q2 | 1.19 (1.00, 1.42) | 0.89 (0.50, 1.58) | 1.28 (0.96, 1.69) |
| BPb Q3 | 1.26 (0.93, 1.71) | 0.70 (0.33, 1.50) | 1.72 (1.09, 2.71)* |
| BPb Q4 | 0.94 (0.65, 1.35) | 0.54 (0.24, 1.20) | 1.20 (0.68, 2.11) |
| Blood Cadmium Quartile | |||
| BCd Q1 | 1.00 | 1.00 | 1.00 |
| BCd Q2 | 1.12 (0.70, 1.77) | 0.83 (0.46, 1.52) | 1.66 (0.72, 3.85) |
| BCd Q3 | 1.12 (0.67, 1.87) | 1.08 (0.65, 1.79) | 1.47 (0.64, 3.39) |
| BCd Q4 | 1.68 (1.12, 2.51)* | 2.59 (1.11, 6.00)* | 1.58 (0.77, 3.24) |
Adjusted for sex, age, race/ethnicity, blood lead levels, obesity, serum cotinine, PIR, smoking status, alcohol consumption, and education level. Quartiles Cadmium: Q1: <0.18 μg/L; Q2: 0.18–0.29 μg/L; Q3: 0.30–0.54 μg/L; Q4: >0.54 μg/L. Quartiles Lead: Q1: <0.70 μg/dL; Q2: 0.70–1.06 μg/dL; Q3: 1.07–1.67 μg/dL; Q4: > 1.67 μg/dL. Tertiles Cotinine: T1: <0.016 ng/mL; T2: 0.016–0.13 ng/mL; T3: >0.13 ng/mL.
indicates p<0.05
indicates p<0.01
indicates p<0.001.
Table 3:
Multivariate logistic regressiona OR (95% CL) of having depression in NHANES 2011–2012 for adult participants (ages 20–47 years)
| All | Men | Women | |
|---|---|---|---|
| N | 1887 | 1007 | 880 |
| Depressed yes, n | 190 | 75 | 115 |
| Depressed no, n | 1697 | 932 | 765 |
| Sex | |||
| Men | 1.00 | ||
| Women | 2.28 (1.63, 3.19)*** | ||
| Age | 1.01 (0.98, 1.03) | 1.00 (0.95, 1.05) | 1.01 (0.97, 1.05) |
| Race/Ethnicity | |||
| Non-Hispanic White | 1.00 | 1.00 | 1.00 |
| Non-Hispanic Black | 1.10 (0.62, 1.95) | 0.65 (0.22, 1.91) | 1.52 (0.69, 3.35) |
| Hispanic (Mexican-American and Other) | 1.20 (0.69, 2.10) | 1.44 (0.66, 3.12) | 1.35 (0.66, 2.78) |
| Other | 1.48 (0.89, 2.46) | 0.87 (0.28, 2.74) | 2.43 (1.00, 5.92) |
| Education level | |||
| Less than High School | 1.00 | 1.00 | 1.00 |
| Completed High School | 1.10 (0.65, 1.87) | 2.00 (1.17, 3.42)* | 0.79 (0.29, 2.13) |
| More than High School | 0.76 (0.55, 1.03) | 1.25 (0.63, 2.49) | 0.61 (0.39, 0.94)* |
| Poverty Income Ratio | |||
| PIR ≤ 1 | 1.00 | 1.00 | 1.00 |
| PIR > 1 | 0.50 (0.30, 0.85)* | 0.40 (0.17, 0.94)* | 0.65 (0.34, 1.24) |
| Body Weight Status | |||
| Underweight/Normal | 1.00 | 1.00 | 1.00 |
| Overweight | 1.47 (0.65, 3.34) | 1.18 (0.42, 3.36) | 1.52 (0.68, 3.42) |
| Obese | 2.30 (1.25, 4.26)* | 1.23 (0.43, 3.54) | 3.36 (1.83, 6.15)*** |
| Alcohol consumption | |||
| No Alcohol | 1.00 | 1.00 | 1.00 |
| 1–4 drinks per week | 0.98 (0.55, 1.75) | 0.49 (0.29, 1.07) | 1.19 (0.52, 2.72) |
| >4 drinks per week | 1.34 (0.60, 2.97) | 0.49 (0.20, 1.23) | 2.76 (0.81, 9.37) |
| Smoking status | |||
| Current Smoker | 1.80 (0.92, 3.53) | 0.76 (0.48, 1.22) | 4.14 (1.17, 14.68)* |
| Former Smoker | 1.54 (0.87, 2.71) | 0.81 (0.37, 1.79) | 2.56 (1.16, 5.67)* |
| Never Smoked | 1.00 | 1.00 | 1.00 |
| Serum Cotinine Tertile | |||
| Serum Cotinine T1 | 1.00 | 1.00 | 1.00 |
| Serum Cotinine T2 | 0.57 (0.31, 1.05) | 0.34 (0.14, 0.82)* | 0.66 (0.30, 1.45) |
| Serum Cotinine T3 | 1.14 (0.63, 2.04) | 1.58 (0.70, 3.60) | 0.82 (0.28, 2.39) |
| Blood Lead Quartile | |||
| BPb Q1 | 1.00 | 1.00 | 1.00 |
| BPb Q2 | 1.19 (0.77, 1.82) | 1.04 (0.44, 2.50) | 1.23 (0.71, 2.13) |
| BPb Q3 | 1.75 (1.12, 2.72)** | 1.51 (0.50, 4.60) | 1.86 (1.01, 3.41)* |
| BPb Q4 | 1.13 (0.61, 2.07) | 0.49 (0.17, 1.42) | 2.97 (1.01, 8.74)* |
| Blood Cadmium Quartile | |||
| BCd Q1 | 1.00 | 1.00 | 1.00 |
| BCd Q2 | 0.97 (0.49, 1.91) | 0.62 (0.28, 1.37) | 1.16 (0.49, 2.74) |
| BCd Q3 | 1.00 (0.36, 2.74) | 1.21 (0.48, 3.05) | 1.05 (0.29, 3.75) |
| BCd Q4 | 1.62 (0.79, 3.32) | 3.16 (1.26, 7.91)* | 1.04 (0.41, 2.62) |
Adjusted for sex, age, race/ethnicity, blood lead levels, obesity, serum cotinine, PIR, smoking status, alcohol consumption, and education level. Quartiles Cadmium: Q1: <0.18 μg/L; Q2: 0.18–0.29 μg/L; Q3: 0.30–0.54 μg/L; Q4: >0.54 μg/L. Quartiles Lead: Q1: <0.70 μg/dL; Q2: 0.70–1.06 μg/dL; Q3: 1.07–1.67 μg/dL; Q4: > 1.67 μg/dL. Tertiles Cotinine: T1: <0.016 ng/mL; T2: 0.016–0.13 ng/mL; T3: >0.13 ng/mL.
indicates p<0.05,
indicates p<0.01
indicates p<0.001.
Table 4:
Multivariate logistic regressiona OR (95% CL) of having depression in NHANES 2011–2012 for adult participants (ages >47 years)
| All | Men | Women | |
|---|---|---|---|
| N | 2016 | 990 | 1026 |
| Depressed yes, n | 239 | 95 | 144 |
| Depressed no, n | 1777 | 895 | 882 |
| Sex | |||
| Men | 1.00 | ||
| Women | 1.44 (0.94, 2.20) | ||
| Age | 0.96 (0.93, 0.98)*** | 0.94 (0.89, 0.99)* | 0.97 (0.94, 0.99)** |
| Race/Ethnicity | |||
| Non-Hispanic White | 1.00 | 1.00 | 1.00 |
| Non-Hispanic Black | 0.87 (0.49, 1.55) | 0.69 (0.27, 1.72) | 0.93 (0.48, 1.80) |
| Hispanic (Mexican-American and Other) | 0.93 (0.44, 2.00) | 0.54 (0.16, 1.79) | 1.35 (0.56, 3.24) |
| Other | 0.58 (0.24, 1.41) | 0.54 (0.26, 1.12) | 0.66 (0.18, 2.42) |
| Education level | |||
| Less than High School | 1.00 | 1.00 | 1.00 |
| Completed High School | 0.69 (0.30, 1.61) | 0.58 (0.14, 2.37) | 0.90 (0.41, 1.95) |
| More than High School | 0.62 (0.20, 1.92) | 0.42 (0.14, 1.21) | 0.96 (0.27, 3.48) |
| Poverty Income Ratio | |||
| PIR1 | 1.00 | 1.00 | 1.00 |
| PIR 1 | 0.31 (0.16, 0.59)** | 0.37 (0.19, 0.72)** | 0.28 (0.12, 0.66)** |
| Body Weight Status | |||
| Underweight/Normal | 1.00 | 1.00 | 1.00 |
| Overweight | 0.78 (0.39, 1.56) | 0.72 (0.27, 1.89) | 0.74 (0.31, 1.77) |
| Obese | 1.13 (0.65, 1.97) | 0.66 (0.25, 1.74) | 1.64 (0.75, 3.55) |
| Alcohol consumption | |||
| No Alcohol | 1.00 | 1.00 | 1.00 |
| 1–4 drinks per week | 1.05 (0.52, 2.15) | 0.62 (0.20, 1.94) | 1.02 (0.43, 2.40) |
| >4 drinks per week | 0.65 (0.15, 2.77) | 0.37 (0.06, 2.41) | 1.04 (0.11, 9.55) |
| Smoking status | |||
| Current Smoker | 1.44 (0.65, 3.20) | 0.73 (0.15, 3.55) | 3.26 (1.43, 7.42)** |
| Former Smoker | 1.39 (0.76, 2.55) | 1.57 (0.47, 5.25) | 1.55 (0.77, 3.14) |
| Never Smoked | 1.00 | 1.00 | 1.00 |
| Serum Cotinine Tertile | |||
| Serum Cotinine T1 | 1.00 | 1.00 | 1.00 |
| Serum Cotinine T2 | 0.93 (0.46, 1.88) | 0.63 (0.21, 1.89) | 1.12 (0.51, 2.45) |
| Serum Cotinine T3 | 0.85 (0.47, 1.52) | 1.07 (0.32, 3.58) | 0.60 (0.24, 1.48) |
| Blood Lead Quartile | |||
| BPb Q1 | 1.00 | 1.00 | 1.00 |
| BPb Q2 | 1.05 (0.45, 2.49) | 0.38 (0.07, 2.21) | 1.49 (0.81, 2.75) |
| BPb Q3 | 0.88 (0.32, 2.38) | 0.15 (0.03, 0.76)* | 1.71 (0.62, 4.73) |
| BPb Q4 | 0.75 (0.28, 1.99) | 0.27 (0.04, 1.72) | 0.94 (0.39, 2.25) |
| Blood Cadmium Quartile | |||
| BCd Q1 | 1.00 | 1.00 | 1.00 |
| BCd Q2 | 1.19 (0.49, 2.87) | 0.83 (0.22, 3.19) | 2.89 (0.75, 11.05) |
| BCd Q3 | 1.19 (0.58, 2.44) | 0.79 (0.40, 1.59) | 2.68 (0.87, 8.24) |
| BCd Q4 | 1.81 (0.73, 4.46) | 1.81 (0.41, 7.99) | 3.30 (1.02, 10.69)* |
Adjusted for sex, age, race/ethnicity, blood lead levels, obesity, serum cotinine, PIR, smoking status, alcohol consumption, and education level. Quartiles Cadmium: Q1: <0.18 μg/L; Q2: 0.18–0.29 μg/L; Q3: 0.30–0.54 μg/L; Q4: >0.54 μg/L. Quartiles Lead: Q1: <0.70 μg/dL; Q2: 0.70–1.06 μg/dL; Q3: 1.07–1.67 μg/dL; Q4: > 1.67 μg/dL. Tertiles Cotinine: T1: <0.016 ng/mL; T2: 0.016–0.13 ng/mL; T3: >0.13 ng/mL.
indicates p<0.05
indicates p<0.01
indicates p<0.001
Blood lead and serum cotinine were not associated with depressive symptoms in all participants (Table 2). After stratification by sex, we observed a non-monotonic association between BPb in women; women in the third quartile of BPb have statistically significantly higher odds to have depressive symptoms (OR=1.79; 95% CL: 1.09, 2.71) compared to the referent quartile (Table 2). Interestingly, after further stratification by age group, the highest third and fourth BPb quartiles were statistically significantly associated with higher odds for having depressive symptoms in younger adult women (20–47 years) compared to the referent BPb quartile (Table 3). Although BPb was not statistically significantly associated with depression in male, further stratification by age, we found an inverse association among male individuals older than 47 years between the third quartile of BPb (OR=0.15; 95% CL: 0.15, 0.76) and current depressive symptoms (Table 4). Current smoking status was a predictive risk factor for depressive symptoms in all participants; however, this association remained statistically significant only in female participants (Table 2). After further stratification by age, current smoking status remained a predictive risk factor for having depressive symptoms in both age groups of women (Table 3 and 4). Moreover, in the younger group (20–47 years of age), former smoking was also a predictive risk factor for depressive symptoms (Table 3). Other than the BCd and smoking status differences observed after sex and age stratification, it is interesting to note that obesity was associated with depressive symptoms only in women (Table 2), and specifically only those women in the younger age group (Table 3).
DISCUSSION
We previously reported an association of blood cadmium with depressive symptoms in young adults (20–39 years) using the NHANES 2007–2010 cross-sectional study. In that study we found that participants in the highest BCd quartile had increased odds of having depressive symptoms compared to those participants in the lowest BCd quartile.13 In the present study, we confirmed the association of BCd with depressive symptoms in all adults (≥20 years) using NHANES 2011–2012. Furthermore, we confirmed that the observed association of BCd with depressive symptoms was independent of cotinine and smoking status. Moreover, in this study, we assessed the role of sex as a potential moderator for the association of cadmium and depressive symptoms. Stratification analyses by sex found that BCd was associated with depressive symptoms in all adult men, and more specifically, in men who were 20–47 years of age. There was also in association of BCd with depressive symptoms in women, but it was limited only to those women in the third and fourth highest age quartiles (>48–60 years and >60 years). It is possible that this association is due to reverse causation since the increase in blood cadmium in the older woman may reflect mobilization of cadmium from bone as result of underlying conditions, such as osteoporosis, which generally affect older, postmenopausal women.12
In this study we also found that younger adult women (20–47 years) in the third and fourth BPb quartiles had statistically significantly higher odds for having depressive symptoms compared to the referent BPb quartile. This finding is similar to that reported previously by Bouchard and colleagues15 of an association between BPb and major depression among young adult (20–39 years of age) participants in NHANES 1999–2004. However, that study did not look at whether the association was restricted to one sex. Analyses of adult (≥20 years) participants of NHANES 2005–200616 and analyses of young adult (20–39 years of age) participants in NHANES 2007–201013 did not find any association between BPb and depressive symptoms. These inconsistencies may be due to the lack of sex stratification in the previous studies and/or to residual confounding. Furthermore, we found an inverse association among individual in the third quartile of BPb and current depressive symptoms in male older than 47 years. We don’t know how lead exposure may have a beneficial role and it might represent a spurious finding.
In the present study, the associations of depressive symptoms with smoking status and obesity are seen in women but not in men. These associations are consistent with previous findings. Women have higher rates of depression,3 and although smoking is less common in women than in men,8 several studies have reported that depression in women is more commonly associated with smoking than in men.9,10
The link of depression and obesity has been thought to be bidirectional, and several studies found that the association between obesity and depression is limited to women, rather than men.22 A possible explanation of these findings in women may lie in the maladaptive coping behaviors, where depressed individuals engage in unhealthy eating (e.g., binge eating, higher caloric intake, etc.) and smoking behavior to cope with their depression.23,24 The cultural stigma of obesity may be a further reinforcing factor with a tendency for obese women to eat in response to negative emotions;23 this could explain the association we observed between obesity and depressive symptoms being confined only to women of reproductive age (20–47 years). Moreover, the association between cigarette smoking and depression is seen to be bidirectional: depression increases the risks of smoking,6,7 and smoking increases the risks of depression.4,5
The underlying biological mechanism of how Cd and Pb may play a role in depression could potentially involve dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis. Cadmium can increase the permeability of the blood brain barrier, leading to intracellular Cd accumulation in the brain in adult rats.25,26 Furthermore, Cd may contribute to the development of depression by perturbing the catecholamine/serotonin system; decreased levels of serotonin, dopamine, and norepinephrine in the brain have been found in adult male rats exposed to Cd.27,28 Lead exposure disrupts catecholaminergic systems; studies in animals show that long-term Pb exposure may result in decreased serotoninergic activity in the brain.29 Impairment of the monoaminergic neurotransmission system is associated with depression and anxiety disorder.2
The use of the structured diagnostic assessments of psychiatric disorders is a strength of the present analysis. The PHQ-9 is widely used in psychiatric research and has a high degree of correspondence with clinical interviews.30 However, the present study also has several limitations, the most important being its cross-sectional design, which limits the inferences that can be made based on the findings. Medical conditions such as osteoporosis, during which Cd and Pb are released from the bone matrix to the blood may have affected our findings. However, we stratified our analyses by age group (20–47 and >47 years of age) in order to minimize the effect that such underlying conditions may have had on our results. As we mentioned previously, the finding in the older female age groups (>47 years of age) of an association between BCd and depressive symptoms may easily be attributable to reverse causation, since mobilization of Cd from the bone may increase BCd levels. The depressive symptoms status in our study is limited to answers in the PHQ-9 about the experience of the participants during the past two weeks, so the depressive symptoms status may reflect a short term health condition, and not necessarily a chronic condition. The half-life of cadmium in blood has been estimated to be 3–4 months,31 whereas lead in adult human blood has been estimated to be from 28 – 36 days.14 Therefore, BCd and BPb, which reflect recent, short-term exposure, may be appropriate to use in this case. The association reported in this study could be biased by uncontrolled factors such as genetic predisposition.2 However, the models were adjusted for several likely important confounding factors. Reverse causation cannot be ruled out as a possible reason for the results since depression may lead to behavioral changes that increase exposure to Cd or Pb (i.e. tobacco smoke). However, the lack of association between cotinine and current smoking status with depression in younger adult men (ages 20–47 years) where we observed the association between Cd and depression suggests that the association we observed may be more than just coincidence. By contrast, the association that we found between BPb and depressive symptoms in young women (aged 20–47 years) may be a result of reverse causation, since in this group we found that tobacco smoking (both current and former users) was also associated with depression, and tobacco smoke is a source of environmental Pb exposure.14
CONCLUSION
If Cd exposure is associated with depressive symptoms and depression, continued efforts at reducing Cd exposure in the general population may help decrease the population incidence of depression. This could be achieved through continued efforts to curb tobacco smoking, since a main source of non-occupational Cd exposure is Cd.12 Additionally, this would have the added benefit of decreasing cadmium exposure from second- and third-hand smoke as well. The finding of an association between BPb and depressive symptoms in young women (20–47 years of age) is intriguing, particularly because we found tobacco smoking (both current and former) as well as obesity to be predictive risk factors to have depressive symptoms in this age group. A recent meta-analysis reported that smoking cessation is associated with reduced depression,32 findings that may help to overcome professional reluctance to intervene with smokers who have mental health problems.33,34 Given that cadmium12 and lead14 are associated with several chronic diseases, the benefits of smoking cessation are multifold by decreasing both the incidence of smoking-related diseases as well as Cd-associated and Pb-associated diseases. However, further studies, such as well-designed prospective studies to evaluate the effect of Cd and Pb exposure on the risk of developing depression are needed to more fully understand the implications of the findings of this study.
Clinical Points.
Cadmium and lead are neurotoxic substances, and few epidemiological studies have addressed the possible association of these metals with depression.
Given that cadmium and lead are associated with several chronic diseases, the benefits of smoking cessation are multifold by decreasing both the incidence of smoking-related diseases as well as cadmium and lead associated diseases.
Footnotes
CONFLICT OF INTEREST: The authors report no conflict of interest.
DISCLAIMER: The findings and conclusion in this report are those of the author and do not necessarily represent the views of CDC/ATSDR.
IRB approval: CDC/ATSDR has determined that our research did not meet the criteria for human research as per federal regulation and therefore did not require review.
REFERENCES
- 1.Stewart WF, Ricci JA, Chee E, Hahn SR, Morganstein D. Cost of lost productive work time among US workers with depression. JAMA 2003; 289: 3135–44. [DOI] [PubMed] [Google Scholar]
- 2.Lanni C, Govoni S, Lucchelli A, Boselli C. Depression and antidepressants: molecular and cellular aspects. Cellular and molecular life sciences. Cell Mol Life Sci 2009; 66: 2985–3008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kessler RC, Berglund P, Demler O, Jin R, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005; 62: 593–602. [DOI] [PubMed] [Google Scholar]
- 4.Weaver TL, Etzel JC. (2003). Smoking patterns, symptoms of PTSD and depression: preliminary findings from a sample of severely battered women. Addict Behav 2003; 28: 1665–79. [DOI] [PubMed] [Google Scholar]
- 5.Aubin HJ, Rollema H, Svensson TH, Winterer G. Smoking, quitting, and psychiatric disease: A review. Neurosci Biobehav Rev 2012; 36: 271–84. [DOI] [PubMed] [Google Scholar]
- 6.Lerman C, Caporaso N, Main D, et al. Depression and self-medication with nicotine: the modifying influence of the dopamine D4 receptor gene. Health Psychol 1998; 17: 56–62. [DOI] [PubMed] [Google Scholar]
- 7.Crone MR, Reijneveld SA. The association of behavioural and emotional problems with tobacco use in adolescence. Addict Behav 2007; 32: 1692–8. [DOI] [PubMed] [Google Scholar]
- 8.Grant BF, Hasin DS, Chou SP, Stinson FS, Dawson DA. Nicotine dependence and psychiatric disorders in the United States. Arch Gen Psychiatry 2004; 61: 1107–15. [DOI] [PubMed] [Google Scholar]
- 9.McKee SA, Maciejewski PK, Falba T, Mazure CM. Sex differences in the effects of stressful life events on changes in smoking status. Addiction 2003; 98: 847–55. [DOI] [PubMed] [Google Scholar]
- 10.Husky MM, Mazure CM, Paliwal P, McKee SA. Gender differences in the comorbidity of smoking behavior and major depression. Drug Alcohol Depend 2008; 93: 176–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Rodgman A, Perfetti TA. The Chemical Components of Tobacco and Tobacco Smoke. CRC Press, Taylor & Francis Group. 2008. [Google Scholar]
- 12.ATSDR. Toxicological profile for cadmium. US Agency for Toxic Substances and Disease Registry, Atlanta, GA: 2012. [PubMed] [Google Scholar]
- 13.Scinicariello F, Buser MC. Blood Cadmium and Depression in Young Adults (20–39 years). Psychol Med 2015; 45: 807–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.ATSDR. Toxicological profile for lead. US Agency for Toxic Substances and Disease Registry, Atlanta, GA: 2007. [PubMed] [Google Scholar]
- 15.Bouchard MF, Bellinger DC, Weuve J, et al. Blood lead levels and major depressive disorder, panic disorder, and generalized anxiety disorder in US young adults. Arch Gen Psychiatry 2009; 66: 1313–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Golub NI, Winters PC, van Wijngaarden E. A population-based study of blood lead levels in relation to depression in the United States. Int Ach Occup Environ Health 2010; 83: 771–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johnson CL, Paulose-Ram R, Ogden CL, et al. National health and nutrition examination survey: analytic guidelines, 1999–2010. Vital and health statistics Series 2, Data evaluation and methods research 2013; 161: 1–24. [PubMed] [Google Scholar]
- 18.Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wittayanukorn S, Qian J, Hansen RA. Prevalence of depressive symptoms and predictors of treatment among U.S. adults from 2005 to 2010. Gen Hosp Psychiatry 2014; 36: 330–336. [DOI] [PubMed] [Google Scholar]
- 20.Pratt LA, Brody DJ. Depression in the U.S. household population, 2009–2012. NCHS Data Brief 2014; 172: 1–8. [PubMed] [Google Scholar]
- 21.Ahlin J, Hallgren M, Ojehagen A, Kallmen H, Forsell Y. Adults with mild to moderate depression exhibit more alcohol related problems compared to the general adult population: a cross sectional study. BMC Public Health 2015;15:542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Carpenter KM, Hasin DS, Allison DB, Faith MS. Relationships between obesity and DSM-IV major depressive disorder, suicide ideation, and suicide attempts: results from a general population study. Am J Public Health 2000; 90: 251–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Musante GJ, Costanzo PR, Friedman KE. The comorbidity of depression and eating dysregulation processes in a diet-seeking obese population: a matter of gender specificity. Int J Eat Disord 1998; 23: 65–75. [DOI] [PubMed] [Google Scholar]
- 24.Leventhal AM, Mickens L, Dunton GF, Sussman S, Riggs NR, Pentz MA. Tobacco use moderates the association between major depression and obesity. Health Psychol 2010; 29: 521–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Mendez-Armenta M, Rios C. Cadmium neurotoxicity. Environ Toxicol Pharmacol 2007; 23: 350–8. [DOI] [PubMed] [Google Scholar]
- 26.Gonçalves JF, Fiorenza AM, Spanevello RM, et al. N-acetylcysteine prevents memory deficits, the decrease in acetylcholinesterase activity and oxidative stress in rats exposed to cadmium. Chem Biol Interact 2010; 186: 53–60. [DOI] [PubMed] [Google Scholar]
- 27.Lafuente A, Márquez N, Perez-Lorenzo M, Pazo D, Esquifino AI. Cadmium effects on hypothalamic-pituitary-testicular axis in male rats. Exp Biol Med 2001; 226: 605–11. [DOI] [PubMed] [Google Scholar]
- 28.Lafuente A, Gonzalez-Carracedo A, Romero A, Esquifino A. Effect of cadmium on 24-h variations in hypothalamic dopamine and serotonin metabolism in adult male rats. Exp Brain Res 2003;149: 200–6. [DOI] [PubMed] [Google Scholar]
- 29.Kala SV, Jadhav AL. Region-specific alterations in dopamine and serotonin metabolism in brains of rats exposed to low levels of lead. Neurotoxicology 1995; 16: 297–308. [PubMed] [Google Scholar]
- 30.Martin A, Rief W, Klaiberg A, Braehler E. Validity of the brief patient health questionnaire mood scale (PHQ-9) in the general population. Gen Hosp Psychiatry 2006; 28: 71–7. [DOI] [PubMed] [Google Scholar]
- 31.Jarup L, Rogenfelt A, Elinder CG, Nogawa K, Kjellstrom T. Biological half-time of cadmium in the blood of workers after cessation of exposure. Scand J Work Environ Health 1983; 9: 327–331. [DOI] [PubMed] [Google Scholar]
- 32.Taylor G, McNeill A, Girling A, Farley A, Lindson-Hawley N, Aveyard P. Change in mental health after smoking cessation: systematic review and meta-analysis. BMJ 2014; 348: g1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Johnson JL, Moffat BM, Malchy LA. In the shadow of a new smoke free policy: A discourse analysis of health care providers’ engagement in tobacco control in community mental health. Int J Ment Health Syst 2010; 4: 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chang CK, Hayes RD, Perera G, et al. Life expectancy at birth for people with serious mental illness and other major disorders from a secondary mental health care case register in London. PloS one 2011; 6: e19590. [DOI] [PMC free article] [PubMed] [Google Scholar]