Abstract: Smoking exacts a huge toll on the health of everyone who becomes addicted to nicotine. That toll is greater in people with HIV, and not only because as a group they smoke more than the general population. Research shows that smoking subtracts more years from the life of a middle-aged person with HIV than HIV itself. Nicotine addiction raises the risk of allcause mortality more in HIV-positive smokers than in HIV-negative smokers. A nationwide study in Denmark found that HIV-positive smokers run a 6 times higher risk of myocardial infarction than HIV-positive never-smokers, but current smoking only doubled the MI risk in HIV-negative people. Research in HIV-positive and negative men and women in the United States found that HIV-positive smokers who had prior AIDS pneumonia had a 3.5-fold higher risk of lung cancer than HIV-negative smokers. Metaanalysis of 18 studies involving more than 625,000 people with HIV found significantly higher rates of smoking-related cancer and infection-related cancer in HIV-positive people than in the general population. People who quit smoking enjoy rapid declines in their risk of smoking-related disease. In the US general population, people who quit by age 40 cut their risk of smoking-related death by 90%. Analysis of almost 5500 SMART trial participants determined that current HIV-positive smokers had a higher risk of five outcomes than former smokers with HIV: all-cause mortality, AIDS-related disease, major cardiovascular disease, non-AIDS cancer, and bacterial pneumonia. In a large DAD study analysis, cardiovascular disease incidence fell steadily as time since quitting rose. A CDC study of a nationally representative HIV population identified numerous factors independently linked to higher smoking prevalence: younger age, white or black race versus Hispanic ethnicity, less education, incarceration, poverty, noninjection drug use, binge alcohol drinking, major depression, and a viral load above 200 copies/mL.
Smoking leads all preventable causes of death across the world.1 And it's more deadly in people with HIV than in HIV-negative people.2 That's not a surprise, because smoking causes deadly diseases -- like myocardial infarction3 and lung cancer4 -- more in people with than without HIV. In North Americans and Europeans with HIV, smoking subtracts more years of life than HIV itself.5
Health professionals need no convincing that smoking imperils health by ravaging one organ after another, from top (stroke) to bottom (peripheral vascular disease). But it's easy to forget the prodigious scope of this assault. In the general population, smoking chops off at least 10 years of life, the CDC estimates,6 and smokers who quit by age 40 limit their risk of a smoking-related death by 90%. In the United States smoking causes more deaths every year than HIV infection, illegal drugs, alcohol, guns, and automobile accidents combined.7 Smoking has killed more than 10 times as many US citizens as all the wars in the history of this country.7
The news gets worse for people with HIV infection. A 2000-veteran study found that smoking raised the allcause death risk more in HIV-positive veterans than in HIV-negative veterans, and lighter smokers ran a death risk just as high as heavier smokers.2 This Veterans Aging Cohort Study (VACS) analysis involved 1034 HIV-positive veterans matched by age, sex, and study site to 739 veterans without HIV. Through a median followup of 5.3 years, 200 veterans with HIV (19.3%) and 72 without HIV (9.7%) died.
Compared with HIV-positive veterans who never smoked, HIV-positive current smokers had more than a doubled risk of all-cause death (adjusted incidence rate ratio [aIRR] 2.31, 95% confidence interval [CI] 1.53 to 3.49).2 In contrast, HIV-negative current smokers did not have an independently higher death risk than HIV-negative never-smokers (aIRR 1.32, 95% CI 0.67 to 2.61). And HIV-positive former smokers did not run a higher death risk than never-smokers (aIRR 1.29, 95% CI 0.81 to 2.04). All-cause death risk was more than 80% higher in HIV-positive smokers than in HIV-positive never-smokers whether they smoked under 20 pack-years (aIRR 1.82, 95% CI 1.20 to 2.76) or more pack-years (aIRR 1.87, 95% CI 1.21 to 2.89). (Pack-years equal the number of cigarettes smoked daily times the number of years a person has smoked. Twenty pack-years means smoking 1 pack a day for 20 years or a half-pack a day for 40 years. (See smokingpackyears.com.)
Analysis of almost 18,000 HIV-positive people in Europe and the United States figured that smoking shortens life expectancy more than HIV itself in the years after people start antiretroviral therapy.5 In a study population that excluded injection drug users, 60% smoked. Through 4 years of follow-up, smokers had a twice higher death rate than nonsmokers (mortality rate ratio [MRR] 1.94, 95% CI 1.56 to 2.41) and (in a separate analysis) a 70% higher death rate than neversmokers (MRR 1.70, 95% CI 1.23 to 2.34) (Figure 1). But mortality among former smokers proved similar to mortality in never-smokers (MRR 0.92, 95% CI 0.64 to 1.34). Compared with nonsmokers, current smokers had 6 times higher mortality from cardiovascular disease (MRR 6.28, 95% CI 2.19 to 18.0) and almost 3 times higher mortality from non-AIDS cancers (MRR 2.67, 95% CI 1.60 to 4.46).
Death Rate Over 4 Years in HIV-Positive Smokers vs. Nonsmokers
Figure 1. A 17,995-person analysis of HIV-positive people in Europe and the United States determined that smokers had almost a twice higher death rate than nonsmokers through 4 years of follow-up.5 In a separate analysis of current smokers, former smokers, and never smokers, current smokers had a nearly twice higher death rate than never smokers. But the death rate was similar in former smokers and people who never smoked.
The researchers calculated that 35-year-old men with HIV would lose 7.9 life-years if they smoked (95% CI 7.1 to 8.7). In comparison, HIV infection would subtract 5.9 years from a 35-year-old man's life (95% CI 4.9 to 6.9).5
HIV-positive current smokers ran a 6 times higher risk of myocardial infarction (MI) than HIV-positive neversmokers in a nationwide Danish analysis (aIRR 6.06, 95% CI 2.99 to 12.25).3 In contrast, current smoking only doubled the MI risk among HIV-negative people compared with HIV-positive never-smokers (aIRR 2.22, 95% CI 1.44 to 3.44). MI risk in HIV-positive former smokers also proved much lower than in HIV-positive current smokers compared with HIV-positive neversmokers (aIRR 2.64, 95% CI 1.16 to 6.01).
The Danish investigators calculated that 72% of MIs in people with HIV could be attributed to smoking, compared with only 24% of MIs in the HIV-negative comparison group (Figure 2).3 That means 3 of 4 heart attacks in people with HIV could be prevented if these people never smoked. Among current HIV-positive smokers, smoking accounted for 42% of MIs. In other words, 42% of MIs could be prevented if all current HIV-positive smokers quit. If all current smokers in the general population quit, it would prevent only 21% of MIs.
MIs Prevented by Never Smoking or Quitting Smoking
Figure 2. If no HIV-positive people ever smoked, 72% of myocardial infarctions would be prevented in people with HIV, according to results of a nationwide Danish study.3 If all current HIV-positive smokers quit, 42% of MIs would be prevented. Proportions of preventable MIs would be much lower in a comparison group of HIV-negative people.
As with heart attacks, so with lung cancer: Your chances of getting it run higher if you smoke and have HIV than if you smoke and don't have HIV.4 That finding comes from a comparison of 2546 HIV-positive and negative women smokers in the Women's Interagency HIV Study (WIHS) and 4274 HIV-positive and negative men smokers in the Multicenter AIDS Cohort Study (MACS). Compared with HIV-negative smokers, HIV-positive smokers who had AIDS pneumonia in the past had more than a 3.5-fold higher risk of lung cancer (aIRR 3.56, 95% CI 1.67 to 7.61). Heavier smoking boosted lung cancer risk in women and men with and without HIV. Compared with people who smoked fewer than 10 pack-years, those who smoked 10 to 30 pack-years had almost a 5-fold higher risk of lung cancer (aIRR 4.75, 95% CI 1.62 to 13.96), and those who smoked more than 30 pack-years had an 11-fold higher risk (aIRR 11.09, 95% CI 3.72 to 33.11).
Rigorous research from around the world shows how thoroughly smoking assails the already-vulnerable health of people with HIV -- in lung, larynx, bone, bladder, and beyond.
- Current smoking more than doubled the risk of bacterial pneumonia in a 5-year study of 4942 people starting antiretroviral therapy in Italy at a median age of 36 (adjusted hazard ratio [aHR] 2.623, 95% CI 2.060 to 3.3339, P < 0.0001).8
- Current smoking almost doubled the risk of bacterial pneumonia over 16 months in 5472 SMART trial participants in 33 countries (aHR 1.82, 95% CI 1.09 to 3.04, P = 0.02).9 Smoking in the past did not independently boost pneumonia risk (aHR 1.64, 95% CI 0.94 to 2.86, P = 0.08). This study group had a median age of 43, and all entered the trial with a CD4 count above 350 cells/mm3.
- Meta-analysis of 14 cohort or case-control studies in HIV-positive adults determined that current smokers had more than a 70% higher risk of bacterial pneumonia than current nonsmokers (HR 1.73, 95% CI 1.44 to 2.06) and almost a 40% higher risk than former smokers (HR 1.37, 95% CI 1.06 to 1.78).10 Former smokers did not run a higher pneumonia risk than never-smokers (HR 1.24, 95% CI 0.96 to 1.60).
- In a 176-person comparison of people with and without HIV in Italy, current smoking was the only independent predictor of chronic obstructive pulmonary disease (COPD).11 HIV-positive participants had a significantly higher COPD prevalence than HIV-negative controls matched for age, sex, and smoking status.
- In the young (median age 36), antiretroviral-naive, international START trial population -- all with a CD4 count above 500 -- every 10 pack-years of smoking independently raised the risk of COPD.12
- A cross-sectional study of 1446 HIV patients in Italy who had thoracic computed tomography found that 48% had emphysema, bronchiolitis, or both.13 Smoking more than tripled the odds of either diagnosis (adjusted odds ratio [aOR] 3.48, 95% CI 2.58 to 4.71) and proved a stronger predictor than age, injection drug use, body mass index, leukocyte count, or gender.
Invasive Pneumococcal Disease
- A nationwide 1995-2012 study in Denmark determined that smoking independently boosted the risk of invasive pneumococcal disease by one third in people with HIV (adjusted relative risk [aRR] 1.34, 95% CI 1.26 to 1.42).14
- Comparing rates of non-AIDS cancers in people with HIV and the general population, metaanalysis of 18 studies involving 625,716 people with HIV found higher rates of smoking-related cancers (lung standardized incidence ratio [SIR] 2.6, 95% CI 2.1 to 3.1; kidney SIR 1.7, 95% CI 1.3 to 2.2; laryngeal SIR 1.5, 95% CI 1.1 to 2.0) and infectionrelated cancer (anal SIR 28, 95% CI 21 to 35; liver SIR 5.6, 95% CI 4.0 to 7.7; Hodgkin lymphoma SIR 11, 95% CI 8.8 to 15) in people with HIV.15
- Smoking doubled the risk of non-AIDS cancers (anal, basal cell carcinoma, Hodgkin lymphoma, lung cancer) in a 3.8-year analysis of 3158 HIV-positive people starting antiretroviral therapy in AIDS Clinical Trials Group (ACTG) studies at a median age of 37 years (aRR 2.12, 95% CI 1.1 to 4.08).16
- Compared with never smoking, current smoking and former smoking more than doubled the risk of death from non-AIDS cancer in a 33,308-person 3-year DAD cohort analysis (aRR 2.20, 95% CI 1.48 to 3.26 for current smoking; aRR 2.52, 95% CI 1.73 to 3.67 for former smoking).17
- Comparison of 3503 HIV-positive people in Denmark and 12,979 people in the general population found almost a tripled risk of smoking-related cancers in the HIV group (aIRR 2.8, 95% CI 1.6 to 4.9) and an 11.5 time higher risk of virus-related cancers (aIRR 11.5, 95% CI 6.5 to 20.5).18 HIV-positive nonsmokers did not run a higher risk of nonvirus-related cancers than general population controls (aIRR 1.2, 95% CI 0.7 to 2.1).
- Compared with HIV-positive people who never smoked, current smokers had nearly a doubled risk of non-AIDS cancer in a 5472-person 33-month SMART trial analysis (aHR 1.8, 95% CI 1.2 to 2.8, P = 0.008).19 Compared with former smoking, current smoking more than doubled the risk of non-AIDS cancer (aHR 2.3, 95% CI 1.5 to 3.6, P < 0.001).
- In 2010 lung cancer displaced non-Hodgkin lymphoma as the leading cancer cause of death among people with HIV in a nationwide French study.20 Among 728 deaths, lung cancer accounted for 61 (8.4%), non-Hodgkin lymphoma for 53 (7.3%), and hepatitis-related cancer for 31 (4.3%).
- Current smoking upped the odds of lung cancer in a Swiss HIV Cohort Study comparison of 68 HIV-positive people with lung cancer and 337 matched controls without lung cancer (aOR for current versus never-smokers 14.4, 95% CI 3.36 to 62.1).21 Neither CD4 count nor antiretroviral therapy predicted lung cancer in this analysis.
- In a 6820-person comparison of HIV-positive and negative men and women in MACS and WIHS, HIV-positive smokers with AIDS pneumonia in the past had more than a 3 times higher risk of lung cancer than HIV-negative smokers (aIRR 3.56, 95% CI 1.67 to 7.61).4
- Mortality after lung cancer proved more than twice higher in HIV patients than in matched HIV-negative controls in a nationwide Danish study (MRR 2.33, 95% CI 1.51 to 3.61).22 All 29 HIV-positive people diagnosed with lung cancer were current or former smokers.
- In a matched comparison of HIV-positive and negative adults in California's Kaiser Permanente healthcare system, 5-year cancer-free survival was significantly lower with lung cancer in the HIV group than in the HIV-negative group (10% versus 19%, P = 0.002) but not in comparisons of HIV-positive and negative people with four other non-AIDS cancers -- anal, prostate, colorectal, or Hodgkin lymphoma.23
HPV Infection and Related Cancers
- A study of 1797 US women with HIV linked smoking to higher prevalence and incidence of human papillomavirus (HPV) infection, including more than a 2-fold higher prevalence of cancer-causing HPV-18 (OR 2.45, 95% CI 1.86 to 3.22).24
- Smoking correlated with cumulative HPV detection in an 11-year Women's Interagency HIV Study of 2543 women with HIV and 895 without HIV.25
- In a 30-month comparison of 328 US women with HIV and 325 without HIV, current cigarette smoking doubled the risk of squamous intraepithelial lesions -- an invasive cervical cancer precursor -- in HIV-positive women (aRR 2.1, 95% CI 1.0 to 4.4, P = 0.06).26
- A study of 2835 US men with and at risk for HIV infection found that smoking independently raised chances of HPV-related external genital warts (aRR 1.2, 95% CI 1.0 to 1.4).27
- A case-control study of Swiss HIV Cohort Study members (85% men) with or without anal cancer found that current smoking more than doubled the odds of anal cancer (aOR 2.59, 95% CI 1.25 to 5.34).28
- A cross-sectional study of 500 Australian men who have sex with men (half with HIV) determined that current smoking doubled the odds of oral HPV (aOR 2.2, 95% CI 1.2 to 3.9), a cause of oropharyngeal squamous cell carcinoma.29 Men with HIV had a significantly higher prevalence of any oral HPV type (19% versus 7%, P < 0.001).
- A 2-year study of US women and men with and at risk for HIV infection found that HIV infection independently raised the risk of incident oral HPV infection, while current smoking increased the risk of oral HPV persistence.30
- Follow-up of 33,308 HIV-positive DAD cohort members for 3 years linked smoking to a doubled risk of cardiovascular death in current smokers (aRR 1.90, 95% CI 1.29 to 2.80) and former smokers (aRR 1.98, 95% CI 1.36 to 2.88).17
- A 5472-person 33-month SMART trial analysis determined that current smoking doubled the risk of major cardiovascular disease compared with never smoking in people with HIV (aHR 2.0, 95% CI 1.3 to 3.1, P = 0.002).19 Compared with former smoking, current smoking raised the risk of major cardiovascular disease 60% (aHR 1.6, 95% CI 1.1 to 2.4, P = 0.02).
- A comparison of 3251 HIV-positive people in Denmark and 13,004 general-population controls matched for age and gender found that current HIV-positive smokers had almost a tripled risk of myocardial infarction compared with current general-population smokers (aIRR 2.83, 95% CI 1.71 to 4.70).3 Previous HIV-positive smokers had almost a doubled risk compared with previous general-population smokers, but that association lacked statistical significance (aIRR 1.78, 95% CI 0.75 to 4.24). MI risk was equivalent in HIV-positive and general-population never-smokers (aIRR 1.01).
- Smoking and family history of cardiovascular disease -- but not diabetes or hypertension -- independently boosted the odds of acute coronary syndrome in a case-control study of people with HIV infection.31 In people with HIV, more than half of acute coronary syndrome diagnoses could be attributed to smoking (54.35%), while lower proportions could be attributed to diabetes (30.58%) or hypertension (6.57%).
- Among 27,136 HIV-positive current smokers in the DAD cohort study, cardiovascular disease risk dropped from an adjusted incidence rate ratio of 2.32 compared with never-smokers in the first year after stopping to 1.46 after more than 3 years since quitting.32
- A 2000-2004 Women's Interagency HIV Study of 1725 women with HIV and 668 women at risk for HIV identified smoking as an independent predictor of metabolic syndrome in the HIV group (aOR 1.31, 95% CI 1.06 to 1.61) and the HIV-negative group (aOR 2.19, 95% CI 1.46 to 3.26).33
Osteoporosis and Fracture
- Meta-analysis of 13 studies of incident fracture in HIV-positive people identified five studies in which smoking independently predicted any fracture or fragility fractures.34
- Meta-analysis of 15 studies involving bone mineral density or fracture in people with HIV or HIV/HCV found that smoking predicted osteoporosis in univariate analysis in four studies, and smoking predicted fracture in univariate analysis in three studies.35
Chronic and Advanced Kidney Disease
- A case-control study of 75 HIV-positive people admitted to the hospital for chronic kidney disease (CKD) and 461 HIV-positive people admitted without CKD determined that smoking independently tripled the odds of hospital admission for CKD (aOR 3.0, 95% CI 1.4 to 5.6, P = 0.005).36 There was a dose-response relationship between packs smoked per day and CKD risk.
- A 35,192-person DAD study analysis determined that current smoking nearly doubled the risk of advanced CKD or end-stage renal disease compared with never smoking (aIRR 1.79, 95% CI 1.08 to 2.97).37
- Smoking quadrupled the risk of oropharyngeal candidiasis in a cross-sectional US study of 215 people with HIV (OR 4.07, 95% CI 1.18 to 14.08, P = 0.027).38
- A 1995-2000 US study of 631 adults with HIV linked smoking to higher odds of prevalent oral candidiasis (OR 2.5, 95% CI 1.3 to 4.8) and incident oral candidiasis (OR 1.9, 95% CI 1.1 to 3.8).39
- A cross-sectional US study 152 people with HIV identified more smoking pack-years as a risk factor for periodontitis.40
- A US study of 415 people with HIV correlated smoking with higher prevalence of oral lesions (including oral candidiasis, salivary gland enlargement, and oral hairy leukoplakia), independently of CD4 count.41
- The impact of smoking on pregnant women and their neonates is well appreciated. In a 1998-2007 study of 1.6 million women with and without HIV across Florida, HIV infection and smoking independently predicted low birth weight, preterm birth, and infants small for gestational age.42