ORIGINAL ARTICLE

Caffeine drinking, cigarette smoking, and dopaminergicreplacement therapy dose in Parkinson’s disease

Carmen Ojeda-Lopez • Amin Cervantes-Arriaga •

Mayela Rodrıguez-Violante • Teresa Corona

Received: 29 May 2012 / Accepted: 22 August 2012 / Published online: 7 September 2012

� Springer-Verlag 2012

Abstract The objective of this study is to assess the

effect of smoking and caffeine intake in the dosage of

dopaminergic replacement therapy. Patients were recruited

from the movement disorders clinic of the National Insti-

tute of Neurology and Neurosurgery in Mexico City. An

interviewer-administered structured questionnaire was

given to all subjects regarding their smoking and caffeine

drinking habits. Dopaminergic replacement therapy infor-

mation was collected and levodopa, dopamine agonists,

and levodopa equivalent daily doses were calculated. 146

Parkinson’s disease patients (50 % female) were included.

All patients were on antiparkinsonian treatment, with a

mean levodopa equivalent daily dose (LEDD) of

550.2 ± 408. Patients were stratified according to smoking

and caffeine drinking status. 104 (71.2 %) of the patients

were ‘‘never smokers’’, 33 (22.6 %) were ‘‘former smok-

ers’’ and 9 (6.2 %) were ‘‘current smokers’’. 40 (27.4 %)

patients reported no history of caffeine intake, 36 (24.7 %)

were former consumers and 70 (47.9 %) were current

caffeine drinkers. No association between LEDD and

smoking or caffeine intake was found. A weak positive

correlation (r = 0.22, p \ 0.04) was found between the

daily dose of pramipexole and the daily intake of caffeine.

LEDD, levodopa daily dose and dopamine agonist daily

dose were not related to smoking or caffeine intake status.

We found a weak correlation between caffeine daily intake

and pramipexole dose. Further prospective exploration is

needed to address the interaction of concomitant A2A

antagonism induced by caffeine intake and dopaminergic

replacement therapy.

Keywords Caffeine � Smoking � Parkinson’s disease �Levodopa daily dose � Dopaminergic replacement therapy

Introduction

Prior smoking and caffeine or tea drinking has been iden-

tified consistently with reduced risk of developing Par-

kinson’s disease (PD) [1]. A dose–response relationship

has also been reported with both smoking and caffeine,

even when adjusting for gender and other sources of bias as

the association of smoking in coffee drinkers [2, 3]. A

recent meta-analysis of prospective studies confirmed that

caffeine intake was inversely associated with PD but most

importantly, it suggests an independent effect of caffeine

and smoking in relation to PD risk [4].

Nicotine stimulates dopamine release, inhibits mono-

amine oxidase B, prevents glutamate-induced neurotox-

icity, and inhibits free-radical damage [5]. It has been

suggested that nicotine or nicotine agonists may have a role

in neuroprotection, in the treatment of motor symptoms and

in the reduction of levodopa-induced dyskinesia [6].

Caffeine owes its primary effects to antagonistic actions

at the adenosine receptors, mainly the A1 and A2A sub-

types, and to other specific neuroprotection signaling

C. Ojeda-Lopez (&)

Department of Neurology, Instituto Nacional de Neurologıa y

Neurocirugıa, Insurgentes Sur 3877, Col. La Fama, Tlalpan,

14269 Mexico, DF, Mexico

e-mail: krmen_ojeda@yahoo.com.mx

A. Cervantes-Arriaga � M. Rodrıguez-Violante � T. Corona

Clinical Neurodegenerative Disease Research Unit, Instituto

Nacional de Neurologıa y Neurocirugıa, Mexico, Mexico

M. Rodrıguez-Violante

Movement Disorders Clinic, Instituto Nacional de Neurologıa y

Neurocirugıa, Mexico, Mexico

e-mail: mrodriguez@innn.edu.mx

123

Neurol Sci (2013) 34:979–983

DOI 10.1007/s10072-012-1180-0

pathways that prevent apoptotic cell death [7]. The A2A

adenosine receptors are expressed on the striatum along

with dopamine D2 receptors. Hence, A2A and D2 receptors

are involved in acute and long-term plastic changes [8].

Few studies have evaluated the effect of coffee intake in

the progression of PD, finding no association between

caffeine intake and rate of progression of the disease [9,

10]. To our knowledge, none of the studies to date has

evaluated the effect of caffeine and nicotine use on PD

treatment.

The objective of this study was to assess the effect of

smoking and caffeine intake in the dosage of dopaminergic

replacement therapy.

Materials and methods

One-hundred and forty-six PD patients who fulfilled the

United Kingdom Brain Bank Criteria [11] were recruited at

the movement disorder clinic at the National Institute of

Neurology and Neurosurgery in Mexico City. All partici-

pants provided written informed consent according to the

determination of the local institutional review board and

ethics committee.

Patient information was collected by an interviewer-

administered structured questionnaire. Demographic and

clinical data were collected including age, age of onset,

severity of PD in terms of Hoehn & Yahr stage (HY),

current antiparkinsonic treatment, and its total daily dose.

Levodopa equivalent daily dose (LEDD) was calculated by

multiplying the total daily dosage of each antiparkinsonian

drug by its potency relative to a standard levodopa prepa-

ration assigned the value of 1 as described elsewhere [12].

Caffeine intake was defined as the consumption of at

least one cup of a caffeine-containing beverage per day or

at least seven cups per week for a period longer than

6 months. Caffeine beverages considered were plain cof-

fee, instant coffee, decaffeinated coffee, espresso, green

tea, black tea and soft drinks (regular and light). Decaf-

feinated beverages were included because most of them are

not actually caffeine-free; another issue is the possibility of

missing a U-shaped relationship because of neglecting the

caffeine content in these drinks.

Patients were classified as ‘‘never drinkers’’, ‘‘former

drinkers,’’ and ‘‘current drinkers’’. Further categorization

was done by dividing patients as ‘‘never drinkers’’ and

‘‘ever drinkers’’.

Information about caffeine beverages intake included

number of cups drank per day. Details of periods of time

with different amounts of intake were recorded and added

up to obtain a single index. To avoid miscalculation of

ounces per cup, a cup of 8 oz and a cup of 12 oz were

shown to the patients. Soft drinks were measured by

number of cans/glasses per day (12 oz). Caffeine cup/year

was calculated by multiplying the average number of cups

per day and the number of years of drinking. To obtain

caffeine daily dose, the concentration of caffeine in the

selected beverages was determined according to the tables

published by Heckman et al. [13]. Patients were also

classified as ‘‘low-dose drinkers’’ (\200 mg/day) and

‘‘high-dose drinkers’’ (200 mg or more).

Tobacco use was classified as ‘‘never smokers’’, ‘‘for-

mer smokers’’ and ‘‘current smokers’’. A second categori-

zation was done by dividing patients as ‘‘never smokers’’

and ‘‘ever smokers’’. The past or current tobacco use was

defined as the consumption of at least one cigarette per day

or at least seven cigarettes per week for a period longer

than 6 months. Smoking pack years (number of cigarettes

per day times the number of years smoked divided by 20)

were calculated. If a patient referred two or more smoking

periods with a significant difference in the number of

cigarettes smoked (double or half), separate pack years

indexes were calculated and added up to create a single

index. Total smoking years were registered. Additionally,

ever smokers were classified as ‘‘moderate smokers’’ (\20

pack years) and ‘‘heavy smokers’’ (20 pack years and

more).

To reduce interview or recall biases, interviewers and

patients were kept unaware of the study hypotheses. The

reliability of the information was verified with the primary

caregiver.

Statistical analysis

To compare groups, we used t tests or Mann–Whitney

U tests for continuous variables. Chi-squared or Fisher’s

tests were used for nominal variables. For three or more

groups (smoking and caffeine intake categories) we used

one-way ANOVA with the Bonferroni post hoc test. Cor-

relation analysis was done using Pearson’s or Spearman’s

correlation factor as needed. A significance level of 0.05

was used throughout. All statistical analyses of data were

performed with SPSS version 16.0.

Results

A total of 146 patients were included, 50 % (n = 73) were

female. Mean age of the sample was 63 ± 11.7 years and

age at diagnosis was 55.9 ± 12.5 (mean duration of the

disease of 7.1 ± 5.1 years).

Regarding the disease severity, the mean HY stage was

2.2 ± 0.7. A total of 91 (62.3 %) patients had mild disease

(HY 1–2), 51 (35 %) patients had a moderate disease (HY

2.5–3), and 4 (2.7 %) of them had a severe disease (HY

4–5). Mean UPDRS part III score was 22.4 ± 12.8.

980 Neurol Sci (2013) 34:979–983

123

All patients were on antiparkinsonic treatment. 108

(74 %) were receiving levodopa (mean time of exposure

1.74 ± 2.6 years). Of these patients on levodopa, 64

(43.8 %) were on monotherapy. 82 (56.2 %) patients were

on pramipexole (38 as monotherapy and 44 combined with

levodopa). No patients were on another dopamine agonist

(bromocriptine or rotigotine).

The mean LEDD of the sample was 550.2 ± 408 (range

17.8–1945, median of 500). For those taking levodopa, the

mean daily dose was 609.1 ± 345 mg (range 50–1625,

median of 500 mg). Mean dose of pramipexole was

1.6 ± 1 mg (range of 0.25–5, median of 1.5 mg).

Smoking and dopaminergic replacement therapy

About their smoking status, 104 (71.2 %) of the patients

were ‘‘never smokers’’, 33 (22.6 %) were ‘‘former smok-

ers’’ and 9 (6.2 %) were ‘‘current smokers’’. Median pack

years in ever smokers were seven and mean smoking years

were 24.3 ± 12.4. Only 23.8 % were classified as heavy

smokers. Table 1 shows the demographic and clinical

characteristics by smoking status along with dopaminergic

replacement therapy data. No statistically significant dif-

ferences were found in clinical variables such as HY stage

(p = 0.06), PD duration (p = 0.20) and UPDRS

(p = 0.75) score between all smoking groups.

No statistically significant difference was found when

comparing use of levodopa or use of pramipexole (p = 0.26

and p = 0.33, respectively). No difference was found in

LEDD between smoking groups (p = 0.98). When analyz-

ing the levodopa daily dose according to the smoking status

no difference was found (p = 0.37); also there was no dif-

ference in pramipexole dose (p = 0.92). No differences

were found when analyzing patients as ‘‘never smokers’’ and

‘‘ever smokers’’; neither when comparing ‘‘moderate

smokers’’ and ‘‘heavy smokers’’. It should be mentioned that

only 14 (25.4 %) of the 55 patients who were current or

former smokers were heavy smokers.

Finally, no correlation was found between pack years or

smoking years with LEDD, levodopa daily dose or pram-

ipexole dose.

Caffeine and dopaminergic replacement therapy

On the other hand, 40 (27.4 %) patients reported no history

of caffeine intake, 36 (24.7 %) were past consumers and 70

(47.9 %) were current caffeine drinkers. Of the 106

patients who were current or former caffeine drinkers, 25

(23.6 %) were classified as high-dose consumers. The

mean caffeine daily intake in ever drinkers was

165.6 ± 141.7 mg. The mean years of caffeine drinking

was 33.8 ± 16.5 (median of 36). Only 8 (6.4 %) patients

began drinking caffeine after PD symptom’s onset.

38 patients (21.5 %) had never smoked nor drank caf-

feine beverages, while only nine (5 %) were current

smokers and caffeine drinkers. Table 2 shows the demo-

graphic, clinical and dopaminergic replacement therapy

data by caffeine drinking status. No statistically significant

differences were found in HY stage (p = 0.53), PD dura-

tion (p = 0.90) and UPDRS score (p = 0.49) among all

caffeine drinking groups.

No statistically significant difference was found when

comparing use of levodopa or use of pramipexole between

caffeine drinking groups. There was no difference in

LEDD (p = 0.29), levodopa daily dose (p = 0.39) or

pramipexole daily dose (p = 0.18).

A low positive correlation (r = 0.22, p = 0.007) was

found between the daily dose of pramipexole and the daily

intake of caffeine. Current and former drinkers consuming

high-doses of caffeine also were on a higher dose of

pramipexole (Mean difference 0.7, 95 % CI of 0.1–1.2,

p = 0.04). No statistically significant correlations were

Table 1 Demographic and

clinical characteristics by

smoking status

HY Hoehn & Yahr stage, LEDDLevodopa equivalent daily dose

Never smokers (n = 104) Former smokers (n = 33) Current smokers (n = 9)

Smoking years – 23.6 ± 13.3 27.1 ± 8.3

Pack year – 19.3 ± 30.6 22.6 ± 30.6

Female gender 61 (58.7 %) 6 (18.2 %) 6 (66.7 %)

Age 63.9 ± 11.5 62.1 ± 12.1 56.1 ± 11.4

Age at onset 56.8 ± 12 55.5 ± 13.3 47.8 ± 13.7

PD duration 7.1 ± 5.3 6.6 ± 4.7 8.3 ± 4.8

HY 2.3 ± 0.7 2.1 ± 0.8 1.7 ± 0.6

LEDD 553 ± 414.1 546.3 ± 342.3 531.3 ± 582

On levodopa 76 (73.1 %) 27 (81.8 %) 5 (55.6 %)

Levodopa daily dose 611.3 ± 340.8 566.7 ± 301 805 ± 609.4

On pramipexole 61 (58.7 %) 15 (45.5 %) 6 (66.7 %)

Pramipexole daily dose 1.6 ± 1 1.6 ± 1.1 1.4 ± 1.3

Neurol Sci (2013) 34:979–983 981

123

found between daily caffeine intake, caffeine drinking

years, and LEDD or levodopa daily dose.

Discussion

Both, tobacco use and caffeine intake, have been impli-

cated with PD risk. Nevertheless, its relationship with the

dopaminergic replacement therapy dosage has not been

addressed. Several studies have evaluated the effect of

nicotine and a selective nicotine agonist in PD symptoms,

some of them obtaining improvements in motor function

and UPDRS scale [14], while others fail to find an anti-

parkinsonic effect [15]. Caffeine and other adenosine A2A

receptors antagonists have shown motor improvement in

PD [16]. On this basis, it could be hypothesized that

smokers and caffeine drinkers might require lower doses of

dopaminergic replacement therapy. To our knowledge, no

previous study has specifically addressed this issue.

We carried out a cross-sectional study of 146 PD

patients to assess the effect of cigarette smoking and caf-

feine intake on the daily dose of levodopa, dopamine

agonists, and LEDD. Our data confirm previous observa-

tions of a low prevalence of cigarette smoking among PD

patients.

Our findings did not show any association between

smoking and caffeine intake with LEDD, levodopa daily

dose, or dopamine agonist dose (pramipexole). Even

though there is evidence that caffeine modifies the phar-

macokinetics and pharmacodynamics of levodopa, short-

ening the latency to motor response (tapping and walking),

and increasing the magnitude of the walking response in

44 % [17]; past studies on chronic administration of caf-

feine showed no antiparkinsonic enhancing effect when

given concomitantly with levodopa, piribedil, or bromo-

criptine [18].

Reports on the effects of smoking and nicotine effects

on PD motor symptoms show contradictory results [6]. In

animal models, caffeine elicits striatal gene expression that

may correlate with biphasic motor responses induced by

different doses of caffeine [19]. Furthermore, it is known

that higher doses of caffeine elicit locomotor depression

through antagonism of A1 receptors [20].

The lack of an association between smoking and caf-

feine intake and LEDD, could also be related to the pres-

ence of daily variations of tobacco and caffeine intake that

do not reflect on LEDD dose. Reports from other authors

[9, 10] show that after the onset of PD symptoms there is

no modification on symptoms progression and caffeine

intake.

Finally, the possibility of an increase in smoking and

caffeine intake due to impulse control disorder (ICD)

associated with dopaminergic drugs cannot be ruled out.

Although the presence of ICD was not assessed in our

study, it seems unlikely since only 7.1 % of the former or

current smokers and 6.4 % of the former or current caffeine

drinkers began consuming after the motor symptoms onset.

A positive but weak correlation was found between the

amount of caffeine intake and the required dose of pram-

ipexole. Pramipexole binds with higher affinity to D3,

which shows heteromerization with both dopamine recep-

tor subtypes and adenosine receptors [8]. We can hypoth-

esize that higher caffeine consumption may alter the

expected modulation, affinity or signaling of the hetero-

meric A2A/D3 receptor complexes thus interfering with

pramipexole. Conversely, an increase in caffeine intake

due to ICD associated with dopamine agonist may be

possible but unlikely as previously discussed.

Table 2 Demographic and

clinical characteristics by

caffeine drinking status

HY Hoehn & Yahr stage, LEDDLevodopa equivalent daily dose

Never drinkers (n = 40) Former drinkers (n = 36) Current drinkers (n = 70)

Drinking years – 31 ± 14.9 35.2 ± 17.2

Caffeine intake (mg) – 178.1 ± 142.3 159.2 ± 141.6

Female gender 20 (50 %) 20 (44.4 %) 37 (52.9 %)

Age 62.5 ± 12 63.7 ± 11.5 62.8 ± 11.7

Age at onset 55.6 ± 11.6 56.4 ± 12.5 55.8 ± 13.2

PD duration 6.9 ± 4.1 7.3 ± 6.3 7 ± 5

HY 2.1 ± 0.6 2.3 ± 0.6 2.2 ± 0.8

Never smoker 30 (75 %) 27 (75 %) 47 (67.1 %)

Former smoker 8 (20 %) 9 (25 %) 16 (22.9 %)

Current smoker 2 (5 %) 0 (0 %) 7 (10 %)

LEDD 493.8 ± 399 504.6 ± 386.4 605.9 ± 421.8

On levodopa 28 (70 %) 26 (72.2 %) 54 (77.1 %)

Levodopa daily dose 415.6 ± 404.1 411.5 ± 413.3 505 ± 393.1

On pramipexole 21 (52.5 %) 20 (55.6 %) 41 (58.5 %)

Pramipexole daily dose 0.6 ± 0.75 1 ± 1 1 ± 1.3

982 Neurol Sci (2013) 34:979–983

123

Limitations of our study include the fact that the pop-

ulation studied was hospital-based and may not reflect the

actual smoking and drinking habits of general population.

Our study is susceptible to potential bias, mainly recall bias

and information bias due to the high variability of caffeine

concentrations in caffeine-containing beverages. One last

concern is the study design, being a cross-sectional study

it cannot differentiate cause and effect from simple

association.

In conclusion, no association was found between

smoking and dopaminergic replacement therapy. Levodopa

daily dose and LEDD were not related with caffeine intake

or smoking habits. An association between high caffeine

daily dose (200 or more milligrams) and pramipexole dose

was found. Although further experimental and clinical

exploration is obviously needed, our findings may suggest

a potential effect of concomitant A2A antagonism induced

by high caffeine intake in patients receiving pramipexole.

Conflicts of interest The authors have no conflicts of interest to

declare in relation to this study.

References

1. Ayuso-Peralta L, Jimenez-Jimenez FJ, Cabrera-Valdivia F, Mo-

lina J, Javier MR, Almazan J (1997) Premorbid dietetic habits

and risk for Parkinson’s disease. Parkinsonism Relat Disord

3:55–61

2. Benedetti MD, Bowe JH, Maraganore DM, McDonnell SK,

Peterson BJ, Ahlskog JE et al (2000) Smoking, alcohol, and

coffee consumption preceding Parkinson’s disease: a case–con-

trol study. Neurology 55:1350–1358

3. Ross GW, Abbot RD, Petrovitch H, Morens DM, Grandinetti A,

Tung KH et al (2000) Association of coffee and caffeine intake

with the risk of Parkinson’s disease. JAMA 283:2674–2679

4. Liu R, Guo X, Park Y, Huang X, Sinha R, Freedman ND, Hol-

lenbeck AR, Blair A, Chen H (2012) Caffeine intake, smoking

and risk of Parkinson disease in men and women. Am J Epi-

demiol 175(11):1200–1207

5. Quik M, Jeyarasasingam G (2000) Nicotinic receptors and Par-

kinson’s disease. Eur J Pharmacol 393:223–230

6. Quik M, O’Leary K, Tanner CM (2008) Nicotine and Parkinson0sdisease: implications for therapy. Mov Disord 23:1641–1652

7. Nakaso K, Ito S, Nakashima K (2008) Caffeine activates de

PI3K/Akt pathway and prevents apoptotic cell death in a Par-

kinson’s disease model of SH-SY5Y cells. Neurosci Lett

432:146–150

8. Ferre S, Ciruela F, Canals M, Marcellino D, Burgueno J, Casado

V et al (2004) Adenosine A2A-dopamine D2 receptor–receptor

heteromers. Targets for neuro-psychyatric disorders. Parkinson-

ism Relat Disord 10:265–271

9. Schwarzschild MA, Chen JF, Tennis M, Messing S, Kamp C,

Ascherio A et al (2003) Relating caffeine consumption to Par-

kinson’s disease progression and dyskinesias development. Mov

Disord 18:1082–1083

10. Simon DK, Swearingen CJ, Hauser RA et al (2008) Caffeine and

progression of Parkinson’s disease. Clin Neuropharmacol

31:189–196

11. Gibb WR, Lees AJ (1988) The relevance of the Lewy body to the

pathogenesis of idiopathic Parkinson’s disease. J Neurol Neuro-

surg Psychiatry 51:745–752

12. Cervantes-Arriaga A, Rodrıguez-Violante M, Villar-Velarde A,

Corona T (2009) Calculo de unidades de equivalencia de levo-

dopa en enfermedad de Parkinson. Arch Neurocien (Mex)

14:116–119

13. Heckman MA, Weil J, Gonzalez de Mejia E (2010) Caffeine

(1,3,7-trmethylxantine) in foods: a comprehensive review on

consumption, functionality, safety and regulatory matters. J Food

Sci 75:R77–R87

14. Ishikawa A, Miyatake T (1993) Effects of smoking in patients

with early-onset Parkinson’s disease. J Neurol Sci 117:28–32

15. Shoulson I (2006) Randomized placebo-controlled study of the

nicotinic agonist SIB-1508Y in Parkinson disease. Neurology

66:408–410

16. Schwarzschild MA, Chen JF, Ascherio A (2002) Caffeinated

clues and the promise of adenosine A(2A) antagonists in PD.

Neurology 23 58(8):1154–1160

17. Deleu D, Jacob P, Chand P, Sarre S, Colwell A (2006) Effects of

caffeine on levodopa pharmacokinetics and pharmacodynamics

in Parkinson disease. Neurology 67:897–899

18. Shoulson I, Chase T (1975) Caffeine and the antiparkinsonian

response to levodopa or piribedil. Neurology 25:722–724

19. Le Foll B, Goldberg SR, Sokoloff P (2005) The dopamine D3

receptor and drug dependence: effects on reward and beyond?

Neuropharmacology 49:525–541

20. Yu L, Coelho J, Zhang X, Fu Y, Tillman A, Karaoz U et al (2009)

Uncovering multiple molecular targets for caffeine using a drug

target validation strategy combining A2A receptor knockout mice

with microarray profiling. Physiol Genom 37:199–210

Neurol Sci (2013) 34:979–983 983

123