Opium Use and Mortality: A Prospective Cohort of 50,000 Adults in Iran

1
Opium Use and Mortality A Prospective Cohort of
50,000 Adults in Iran

• Farin Kamangar

2
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

3
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

4
Opium and Opioid Use

• Recreational and medicinal use
• Long history
• 13 22 million used opium or its derivatives in
  2008
• Many areas of the world

5
Opium Recreational Use (19th Century)
6
Opium Medicinal Use
7
Opium Distributed and Used Worldwide
8
Long History Goddess of Opium
9
Tintin and Opium Use
10
Epidemiologic Studies of Opium and Health

• Very few epidemiologic studies of chronic opium
  use, despite such widespread use and long history
  
• Only a few case-control studies

11
Studies of Opium and Health

• Increases the risk of esophageal cancer
  (Ghadirian 1985, Nasrollahzadeh 2008)
• Increases the risk of bladder cancer (Behmard
  1981, Aliasgari 2004, Shakhssalim 2010, Hosseini
  2010)
• Increases the risk of laryngeal cancer (Mousavi,
  2003)
• Increases the risk of two other cancers
  (unpublished)

12
Studies of Opium and Health

• Increases the risk of cardiovascular disease
  (Sadeghian 2007, Sadeghian 2010, Masoomi 2010,
  Masoumi 2010)
• Increases the risk of deep vein thrombosis
  (Safaii 2010)
• Increases the risk of lung disease with chronic
  obstruction (Da Costa 1971)

13
Research Gap

• No large-scale or long-term prospective cohort
  studies of opium and health

14
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

15
Objectives

• Use data from Golestan Cohort Study
• To determine the effect of opium use on overall
  and cause-specific mortality
• To determine this effect by type of opium use
• Teriak
• Shireh
• Sukhteh
• Heroin

16
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

17
Golestan Cohort

• In Golestan Province, northeast of Iran
• 50,045 adults, 40 to 75 years of age
• Detailed questionnaire data
• General Questionnaire
• Food Frequency Questionnaire
• Biologic samples
• Blood
• Urine
• Hair
• Nail

18
(No Transcript)
19
(No Transcript)
20
Very High Rates of Esophageal Cancer

• Anecdotal reports
• Reports from central institutions (Habibi 1965,
  Haghighi 1971)
• Exploratory observations (Kmet 1966)
• Cancer registry (1968 to 1971, Iran-IARC
  collaborations)

21
(No Transcript)
22
Registry Results

• The highest rates ever reported in the world
• Rates also very high in women
• Large differences in rates in 500 kilometers

23
Ecologic Studies in the 1970s

• Differences in fauna and flora
• Differences in foods consumed (sheeps milk,
  fruit and vegetables, sesame oil, etc)
• Hot tea consumption in high quantities
• Chewing of nass (a mixture of ash, lime, and
  tobacco)
• Opium use (up to 50 of both sexes over 35 years)

24
Case-Control Study(350 cases and 700 controls)

• Not associated sheeps milk, sesame oil, chewing
  of nass, use of pregnancy diets, salting and
  sun-drying
• Moderately associated cigarette smoking,
  drinking hot tea, low intake of fresh fruits and
  vegetables, low socio-economic status
• Not assessed opium use
• 
  Cook-Mozaffari, BJC 1979

25
Opium Further Studies

• Types
• Crude opium
• Sukhteh (opium pipe residues) mutagen in Ames
  test
• Shireh (refined opium extract)
• Hewer , Lancet 1978
• Case-control study of families
• A half-completed case-control study showed a
  non-significant two-fold risk associated with EC
  risk
• Ghadirian, IJC 1985

26
(No Transcript)
27
Summary Results from the 1970s

• Northeastern Iran a very high-risk area for EC
• Risk similar in men and women
• Smoking a weak risk factor, and alcohol
  consumption very uncommon
• Some risk factors suggested (low SES, low intake
  of fresh fruit and vegetables, hot tea intake,
  but none striking)
• Opium some suggestions, but studies incomplete

28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
Case-Control StudyBasic idea

• To re-evaluate the previous hypotheses for the
  very high incidence of esophageal cancer (e.g.,
  opium use, hot tea intake, low socioeconomic
  status, low intake of fresh fruits and
  vegetables)
• To test new hypotheses (e.g., animal contact,
  oral hygiene, genetic studies)
• Use new technology (e.g., new genotyping
  capabilities, proteomics)

32
(No Transcript)
33
Cases and controls

• 300 esophageal cancer cases (all squamous cell)
• 600 neighborhood controls, matched for age, sex,
  and neighborhood of residence

34
(No Transcript)
35
Data and samples

• General questionnaire
• Food Frequency Questionnaire (validated against
  24-hour recalls)
• Blood
• Hair and Nail
• Biopsy samples

36
General Questionnaire

• Demographic characteristics
• Residential history
• Occupational history
• Family and SES
• Personal history of disease
• Upper GI signs and symptoms
• Personal and family history of cancer
• Tobacco and opium use
• Alcohol consumption
• Medication history
• Oral health
• Anthropometric indices
• Animal contact
• Tea consumption
• Physical activity
• Methods of cooking and food preservation and
  water source

37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
Esophageal Cancer Lecture for Turkmen Doctors
42
(No Transcript)
43
(No Transcript)
44
Case-Control Study

• Recruitment started in 2003 and was completed in
  2007.
• A total of 300 cases and 571 matched controls
  were recruited.
• Several papers have been published.

45
Opium and Tobacco Use in Relation to Esophageal
Cancer

• Tobacco and/or opium
• Neither 1.00
• Tobacco only 1.70 (1.05 2.73)
• Opium only 2.12 (1.21 3.74)
• Both 2.35 (1.50 3.67)
• Some 30 of all cases can be attributed to these
  two factors.
• Nasrollahzadeh et al. BJC

46
(No Transcript)
47

• Cohort Study

48
Cohort Study

• This cohort study was designed to enroll 50,000
  subjects from the eastern part of Golestan
  Province, with objectives similar to the
  case-control study.
• Design was very similar to the case-control
  study.
• General questionnaire
• Dietary questionnaire
• Blood
• Hair and nail
• Urine

49
Cohort Study

• Plasma (x8), RBC (x2), and buffy coat (x4) are
  stored in bar-coded straws and kept in 80
  freezers or nitrogen tanks.
• Urine samples (used for PAH and opium assays) are
  also stored in -20 freezers or nitrogen tanks.
• Half the samples in Iran and half in IARC.

50
(No Transcript)
51
Cohort Study

• Outcomes
• Death (from any cause)
• Incident cancer (any site)
• Esophageal and gastric cancers (emphasis)
• Outcome assessment
• All relevant data are collected from hospitals
  and clinics
• Two internists assign codes may be a third one
• For EC and GC an international panel of experts

52
Cohort Study

• Recruitment started in 2004 and was completed in
  May 2008.
• Follow-up continues to date

53
Cohort Study Follow-Up

• Annual follow-up
• Private phone (available to 98 of the
  participants)
• Phone numbers from relatives and friends
• Home visits
• In rural areas the Behvarz (local health person)
• Follow-up through May 2011
• 234,928 person-years
• gt 99 success rate
• 2,145 deaths (130 due to esophageal cancer)

54

• Case-control
• GEMINI Cohort

55
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

56
Regression model

• Cox proportional hazards models to estimate the
  hazard ratios and 95 confidence intervals
• Adjustment for potential confounders
• Assessment of interactions
• Sensitivity analyses

57
Exposure and Outcome Variables

• Exposure Opium use
• Type (teriak, shireh, sukhteh, heroin)
• Typical amount of use in nokhod ( 0.2 g)
• Duration of use
• Frequency of use
• Outcome Death
• Overall
• Cause-specific

58
Other Variables

• Potential confounders and effect modifiers
• Age
• Sex
• Place of residence
• Ethnicity
• Marital Status
• Highest educational level
• Cigarette smoking
• Body mass index
• Height
• Consumption of fruit and vegetables
• Socioeconomic status indicators

59
Pilot Studies

• Reliability and validity of opiate use
  self-report in a population at high risk for
  esophageal cancer in Golestan, Iran. (Cancer
  Epidemiol Biomarkers Prev. 2004)
• Golestan cohort study of oesophageal cancer
  feasibility and first results. (Br J Cancer.
  2005)
• Validity and reliability of a new food frequency
  questionnaire compared to 24 h recalls and
  biochemical measurements pilot phase of Golestan
  cohort study of esophageal cancer. (Eur J Clin
  Nutr. 2006)

60
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

61
Study Population

• 50,045 participants
• 234,928 person-years of follow-up
• Median follow-up 4.7 years
• Mean age at enrollment 52 years
• 58 women
• 74 of Turkmen ethnicity
• 70 no formal education
• 17 smokers
• 17 opium users

62
Opium Distribution Routes
63
Global Opium Consumption
64
(No Transcript)
65
All-Cause Mortality

• HR (95 CI)
• Unadjusted 2.26 (2.06 2.47)
• Adjusted 1.86 (1.68 2.06)
• Almost doubling the risk
• Much stronger than cigarette smoking
• (Khademi, BMJ 2012)

66
Opium and MortalityAdjusted HR (95 CI)

• All Causes (n 2145) 1.86 (1.68 2.06)
• Circulatory (n 1073) 1.81 (1.56 2.09)
• Cancer (n 449) 1.61 (1.28 2.03)
• Respiratory (n 95) 3.78 (2.36 6.04)
• Digestive (n 69) 3.12 (1.82 5.37)
• Infectious (n 51) 5.47 (2.87 10.4)
• External (n 135) 0.86 (0.54 0.35)
• Other (n 150) 1.66 (1.11 2.48)
• Unknown (n 123) 2.42 (1.59 3.67)

67
By Duration of Use(only in those who used it
before diagnosis of major diseases)

• Never users 1.00
• Former users 1.27
• Current users 1.70
• 3 years 1.62
• 4 to 7 years 1.39
• 8 to 12 years 2.00
• 13 to 20 years 1.79
• 21 years 1.90
• 
  p for trend lt 0.001

68
By time since last used
69
Subtype Analyses Sensitivity Analyses

• Teriak, shireh, and heroin all increased risk
• Both smoking and ingesting opium increased risk
• Excluding deaths that occurred in the first 6,
  12, 18, and 24 months of follow-up had little
  effect on the results
• Further adjustment didnt change the results
• Results remained unchanged in a healthy
  subcohort analysis

70
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

71
Evidence for Causality (1)

• Consistent with results of case-control studies
  of cancer and cardiovascular diseases
• Relatively strong association (hazard ratio of
  2), stronger than tobacco use in this population
• Dose-response association with duration of use
• Risk reduction after cessation of exposure

72
Evidence for Causality (2)

• Adjusting for multiple confounders (age, sex,
  ethnic group, place of residence, tobacco
  smoking, education, intake of fruit and
  vegetables, )
• Measures taken to study reverse causality
• Excluding those who started opium use after
  diagnosis of major diseases
• Excluding outcomes within 24 months from the
  initiation of the cohort

73
Evidence for Causality (3)

• Biologic plausibility Much to be learned, but
  alkaloid and non-alkaloid components have a
  number of systemic effects. Examples
• Lung and Bronchi
• Activation of opioid receptors in medulla,
  resulting in reduced breathing frequency and
  tidal volume
• Upper airway constriction
• Cancers
• Activating angiogenesis and neovascularization
• PAHs due to smoking
• Liver
• Perivenular and focal hepatocelluar necrosis

74
Evidence for Causality (4)

• Opium users in this cohort use very low amounts
  of opium (average of 0.6 g/day), are at very low
  risk of overdose, and are NOT very different from
  other study participants for most other studied
  variables.
• Results mostly reflect long-term effects of low
  amounts of use

75
Implications

• Opium and opioids are widely used. Death due to
  opium use is not limited to overdose.
• This study may (and only may) have implications
  for long-term use of opioids in medicine.

76
Outline

• Significance of the Study
• Objectives
• Golestan Cohort Study
• Statistical methods
• Results
• Discussion
• Lessons learned and suggestions for further
  research

77
Further Research

• Case-control studies of various outcomes in
  relation to opium use (keep in mind reverse
  causality)
• Studies of the biologic effects of opium and
  opioid use
• Investigating long-term effects of opioid
  analgesics

78
Lessons Learned

• Just do it!
• There is a story and a rationale behind every
  study but things take you where you dont
  necessarily expect.
• For more cohorts, try to find a few unique
  things.
• Leadership, collaborations, and patience All
  very important.

79
Further Reading

• Khademi et al. Opium use and mortality in
  Golestan Cohort Study. BMJ, 2012
• Pourshams et al. Cohort Profile The Golestan
  Cohort Study. Int J Epidemiol, 2010.
• Kamangar et al. Esophageal cancer in Iran. Arch
  Iran Med, 2007.

80
(No Transcript)
81
(No Transcript)
82
(No Transcript)