You are in Research

Differential effects of medical marijuana based on strain and route of administration: A three-year observational study

Valerie Leveroni Corral
Wo/Men's Alliance for Medical Marijuana
309 Cedar Street #39
Santa Cruz, CA 95060

A collective of patients and caregivers, creating community, building hope, dissolving barriers, providing support and free medical marijuana since 1993


Cannabis displays substantial effectiveness to affect a variety of medical symptoms. Seventy-seven patients took part in a study in California to assess the efficacy of organically grown Cannabis sativa and indica strains in treatment of various medical conditions via smoking or ingestion. HIV/AIDS was the most frequent condition reported, at 51%. Standardized rating forms provided 1892 records that were statistically analyzed. Results demonstrated that in the case of nausea and spasm, changes in symptom expression are definitely affected by method of cannabis administration. However, while Cannabis indica strains increased energy and appetite, it is useful to note that in treating nausea in HIV/AIDS and orthopedic diagnosis groups, Cannabis sativa and C. indica strains proved equivalent.

Keywords: cannabis, medical marijuana, Cannabis sativa, Cannabis indica, AIDS, HIV


Marijuana, whether Cannabis sativa and Cannabis indica, produces its medical and other effects by virtue of the concentration and balance of various active ingredients, especially the cannabinoids, which are unique to marijuana, but including also a wide range of terpenoids and flavonoids (McPartland and Mediavilla 2001; McPartland and Pruitt 1999). Terpenoids are cannabis constituents that provide the characteristic strong odor of marijuana and hashish. Flavonoids are any of the flavone derivatives. The concentration and relative proportions of these ingredients depend on the plant's genetic structure and applied hybridization techniques, and as such, allow for a substantially varied outcome.

Little is known about how differences in constituent profiles translate into differences in therapeutic effectiveness. A range of differentiable effects has been ascribed to THC (tetrahydrocannabinol is the primary psychoactive component of marijuana) and CBD (cannabidiol, a compound related to THC) when administered in purified form (Iversen 2000). Studies are lacking on the differential clinical effects produced when varying "menus" of constituents are taken together.

Another factor bearing on the effects and the effectiveness of marijuana is the route of administration. Orally administered marijuana is absorbed more slowly than when delivered systemically (e.g., smoking, vaporizers). Moreover, the liver metabolizes orally ingested marijuana. This produces a potent and long-acting cannabinoid (11-hydroxy-THC), which induces varied reactions in medical marijuana patients and is not often well tolerated (Grotenhermen 2001). However, once more, there is little information available concerning the differential clinical effects of oral vs. smoked forms of marijuana.

A major obstacle to obtaining data concerning differential clinical effects produced by varying strains of cannabis and by different routes of administration is, of course, the common illegality of medical marijuana use. Almost equally troublesome, however, is the widespread view that medical knowledge can be gained only through randomized controlled trials. It is becoming increasingly accepted that valid causal inferences can be, and frequently are drawn quite regularly in medicine without such studies. As such, observational studies are quite capable of generating useful information, provided due care is taken to keep careful track of the process. In this case, careful and consistent documentation would be required concerning: 1) which forms of marijuana are being taken, by what route, and: 2) what outcome is experienced by patients.

The passage of Proposition 215 in California in 1996 legalized medical marijuana under state law, thus clearing some legal obstacles to research. Prior to the passage of Proposition 215, two or more cannabis buyers’ clubs and our collective comprised of patients and caregivers were in operation. Several provider associations have been operating since that time despite harassment of some by law enforcement agencies.

Valerie Leveroni Corral founded the Wo/Men’s Alliance for Medical Marijuana, WAMM in 1993. WAMM is a collective of patients and caregivers attempting to create community, build hope, dissolve barriers, and provide support and medical marijuana at no cost to patient members who possess a signed and verified recommendation from a physician licensed to practice medicine in California. A genetically monitored, organic, communal garden is tended by WAMM client/ participants under the direction of Mike Corral and Valerie A. Leveroni Corral.

A primary function in this community based educational system is the creation of a database of information regarding the treatment of different symptoms with distinct cannabis varieties. This is achieved through daily effectiveness surveys and statistical analysis. Our present collection of data also includes measures of effectiveness of cannabis on other autoimmune illnesses, such as systemic lupus erythematosis, as well the many other disorders, including muscular dystrophy, epilepsy, quadriplegia, paraplegia, Parkinson's disease, glaucoma, arthritis, fibromyalgia, depression and migraine. However, AIDS and HIV-related conditions are the most frequently represented among our clientele.

WAMM initiated a study in 1993 designed to address the question of differential clinical effects between Cannabis sativa and C. indica strains and hybrids, and also examining effects of inhaled and ingested routes of administration. This study is ongoing and now includes "blind" trials where the varieties used are not apparent to the participating patient. A statistician determined all analyses. [Tables 17 & 18]

Materials and Methods

The determination of the variety of cannabis was based on the country of origin of the seeds strains and physical characteristics of each plant variety. We assure the genetic purity through carefully controlled breeding techniques, substantiated by twenty-five years of experience in cultivation, propagation and breeding of cannabis. Personal interaction took place with patient use of cannabis in more than one hundred different terminal cases.

An assessment instrument form is provided weekly to participating patients [see Tables 17 & 18]. The patient places a label from a weekly supply on the seven day form, denoting the variety and form of cannabis (inhaled or ingested), the number of "puffs" if inhaled medicine is used and the amount or weight employed. All participants were instructed in a specific method for inhaling. Patients were requested to use and denote dosages correlated to the relief of specific symptomatology. Participants observed and rated symptoms before and after cannabis use to record their severity. This is done upon rising from sleep in all cases except "insomnia" and prior to using any cannabis. Assessments were made weekly, at minimum, or as much as seven times per week, in order to assess effectiveness and of different strains upon different target symptoms.

Findings were derived from data gathered during the time period of June of 1993 into early 1997. Statistical analysis consisted of frequency analysis, paired T-tests of "before" and "after" scores on each measured symptom or condition, and a series of one way ANOVAs on route of administration (either inhaled or ingested, cannabis strain, and diagnosis).

Because the therapeutic effects of cannabis are sometimes ascribed to its mood-altering effects, we also performed a correlation analysis of the change in mood score with other outcome variables.

Inhalation methods of cannabis consisted mostly of smoking, with some use of vaporization, although patient reports of effectiveness appear substantially lessened when this technique was employed. This could certainly depend on the quality of the vaporizer design.

Ingested forms of cannabis consisted of baked goods and "mother's milk" (a soymilk-based liquid), and a whole cannabis tincture made with pure grain alcohol with leaf or a combined blend of leaf and flowers. Strains of marijuana were C. sativa and C. indica and their hybrids. The morphological distinction between these strains was determined by experienced cannabis cultivators associated with WAMM, based on characteristic features of the two sub-species, varieties or strains.

These sub-species varied from week to week and included the following pure strains and hybrid strains:

C. sativa, C. indica, as well as hybrids of both, being the identified female C. sativa x male C. indica, as well as the identified female C. indica x male C. sativa. We secured a method of analysis of the chemical content of test materials, although we believe that the findings may be subject to error. Results from a drug detection laboratory indicated that C. sativa measured: THC 23.7%, CBD <0.1% and CBN <0.1%. Results indicated that C. indica strains measured THC 19.6%, CBD <0.2% and CBN <0.5%. Cannabis potency testing results by ElSohly Labs of the same sample of C. sativa after storage for eight months yielded a value of THC 17.6 %.


Seventy-seven patients completed a total of 1892 forms (range 1 - 256, median 8) during the three-year study period. Of these, 43 were male (56 percent), 22 were female (29 percent) and 12 were not coded as to gender. The distribution of primary diagnosis is presented in Table 1. [Table 1]

Thirty-nine patients (51 percent) had HIV/AIDS; 14 (18 percent) had neurological diseases, and 7 (9 percent) had a principle diagnosis of cancer.

To avoid biasing results due to a large proportion of questionnaires being completed by relatively few patients, we standardized the analysis by reviewing a maximum of eight records per patient, the median number completed by study subjects. These records were randomly chosen. Accordingly, our analysis contained 432 records. Of these, 261 (61 percent) referred to C. sativa experiences; 65 (15 percent) were C. indica, while 105 (24 percent) were coded "other". Certain types of marijuana were donated or undeclared, we labeled these as "other" and included them in our findings. Ingested forms were also recorded [Table 4]. Some entries were coded with missing information, entered as slang or incorrectly named, these were excluded.

Paired t-tests of before and after health status revealed that the following symptoms were relieved to a statistically significant extent by marijuana (without regard to strain or route of administration): pain, energy, mood, nausea, appetite, and awareness. The remaining symptoms were not reliably relieved to this extent. Table 5 and Table 6 show the scores on each variable. The magnitude of improvement was unrelated to clinical diagnosis, as determined in ANOVA [Table 10], with one exception: the degree of relief of nausea was greater in the HIV/AIDS group (4.54 units) than in the orthopedic group (1.58 units) to a (marginally) statistically significant extent (p = 0.04).

We next performed ANOVA on the strain of marijuana ingested: C. sativa and C. indica. The mean change scores, "before" scores minus "after" scores for patients with each condition. For the most part, some observed changes were unrelated to strain of marijuana. However, two symptoms - energy and appetite - were improved to a statistically greater extent by C. indica than by either C. sativa or "other."

C. indica produced a mean improvement in energy of 3.76 units (vs. 1.53 for C. sativa and 2.22 for "other") and a mean improvement in appetite by 5.22 units (vs. 3.41 for C. sativa and 4.32 for C. indica). These differences were significant at the 0.012 and 0.005 levels, respectively [Table 8].

ANOVA was then conducted using route of administration as the independent variable [Table 6 & Table 7]. For the most part, ingested and inhaled marijuana had similar magnitudes of effects. Only one symptom - spasm - showed preferential improvement using smoked over ingested marijuana (p = .036). [Table 6]. Patients reporting "other" routes of administration, such as ingestion, had substantially less relief of nausea than patients inhaling or ingesting marijuana [Table 7].

It is reported that THC may reduce spasms associated with both neurological and non-neurological disorders (Hollister, 1986; British Medical Association Report, 1997). It is interesting to note that the non-psychoactive cannabinoid cannabidiol has been shown to exhibit anticonvulsant properties in certain animal studies (Iversen 2000)(The Science of Marijuana, L.L. Iversen, Ph D.) In the case of some patients it has been noted to reduce or prevent the onset of both spasms and seizures when used alone or as an adjunct medicine. It appears that there are receptor sites for cannabinoids that have beneficial effects on seizure activity.

Finally, analysis of the Pearson correlation coefficients between changes in mood scores and changes in other symptom scores revealed only a single statistically significant correlation, between mood and energy level (p = 0.035). Mood was not correlated with any other outcomes, including pain relief (p = 0.817) [Table 11].


We analyzed 432 records of therapeutic cannabis exposures, including information on strain (C. sativa, C. indica, or other), and route of administration (inhaled, ingested or other). The outcome variables consisted of scores to a series of questions on symptoms, completed by the patient both before and after administering cannabis medicines.

Results indicate that cannabis was uniformly effective in relieving symptoms across a wide range of diagnostic categories. No differences were observed in the extent to which symptoms were relieved based on diagnosis, except that patients with HIV/AIDS experienced more relief of nausea than patients with primary orthopedic diagnoses [Table 13].

On several occasions, terminally ill patients remarked upon a recurrent phenomenon, described as a “shift in consciousness" or "perception" allowing them to approach their impending death more "openly" or in a more "relaxed" manner. This is of particular interest, as each patient also reported a reduction in anxiety often associated with the dying process. Future studies will further examine measures anxiety in the cannabis patient population.

C. indica appeared to be superior to C. sativa and "other" in improving energy and appetite [Table 9]; otherwise, no differences in strain effects were observed. Route of administration had little effect on outcome in our series. Two symptoms, spasm [Table 6] and nausea [Table 7] showed preferential improvement of smoking as compared ingestion. In no condition was the ingested route superior to smoking upon symptom management.

Changes in mood were not correlated with changes in other outcomes except for a modest correlation with energy [Table 11]. The finding that mood did not correlate with other outcomes casts doubt on the theory that therapeutic cannabis effects are related primarily to improvement in mood. Conversely, this may have something to do with the notion suggested by some patients that mood is not necessarily correlated to the concept of "feeling better." In our findings, it appeared that mood was often independent of symptom expression. This result is interesting because it appears from written testimony by patients in their surveys that they believe changes in awareness or consciousness do affect overall healing. We plan to further examine the validity of these phenomena in future studies.

These findings support that few differences were noted by patients between C. sativa and C. indica strains and between ingestion vs. inhaled routes of administration. This is likely due to modest observed differences in cannabinoid content in the supplied strains. We hope that a reliable and accessible means of analysis will become available in the near future.

This study is limited by the lack of blinding. For this reason, in 1998 a revised protocol was instituted in which patients receive a one-week supply of therapeutic cannabis at a time without knowledge of particular variety provided. Patients continue completing forms on a weekly basis. This method of blinding is expected to provide a more rigorous test of any distinctions between C. sativa and C. indica strains. Results may have implications for subsequent crossbreeding of strains to maximize therapeutic effects.

This study is only a small first step in the attempt to develop improved cannabis medicines to affected patients. The most significant current limitation to this type of research is the absence of a convenient legal mechanism in the USA for analyzing cannabis samples for biochemical constituent content. Until this limitation is overcome, progress in this area will be slow at best.

On the other hand, we should not underestimate the value of clinical observation in judging strains of cannabis and their differential clinical effects irrespective of chemical content. Thus, while the work we report here does not definitely address issues of chemical variability, we believe that our findings provide at the very least a good working hypotheses for use in future studies.


Grotenhermen, Franjo. 2001. Practical hints. In Cannabis and Cannabinoids: Pharmacology, toxicology and therapeutic potential, edited by F. Grotenhermen and E. Russo. Binghamton, NY: Haworth Press.

Iversen, Leslie L. 2000. The science of marijuana. Oxford ; New York: Oxford University Press.

McPartland, J. M., and P. L. Pruitt. 1999. Side effects of pharmaceuticals not elicited by comparable herbal medicines: the case of tetrahydrocannabinol and marijuana. Altern Ther Health Med 5 (4):57-62.

McPartland, John M., and Vito Mediavilla. 2001. Non-cannabinoids in cannabis. In Cannabis and cannabinoids, edited by F. Grotenhermen and E. B. Russo. Binghamton, NY: Haworth Press.

Received: September 21, 2000

Accepted in Final Form: April 29, 2001.


Purpose of the Project

  • To determine if there are physical, mood and perception changes resulting from use of the test article.
  • To determine if the method of delivery affects measures of effectiveness.
  • To determine if different types of cannabis affect diagnoses and measures of effectiveness.
  • To assess the correlation between changes in mood and other measures of effectiveness.

Summary of Population

N = 77

43 males (56%)

22 females (29%)

12 missing gender distinction (15%)

Table 1

Description of Population by Primary Diagnosis

Table 2

Description of Patient Population by Secondary Diagnosis

Table 3

Questionnaire Structure Measures of Effectiveness




Desired Effect





















Muscle Spasms
























Questionnaire Logistics

  • 1892 Questionnaires Completed over 3 years
  • Range of 1 to 256 questionnaires
  • Average of 8 questionnaires/patient
  • Analysis completed based on the average number of questionnaires completed (to normalize data for analysis)

Table 4

Statistical Methods
432 questionnaires analyzed

Frequency analysis, Paired t-tests, Paired t-test correlations, One Way ANOVA, Post-Hoc (Bonferroni), Pearson Correlation and Multivariate tests performed

One Way ANOVA conducted on variables using the following 3 groups


Group 1 – test article “ingested”


Mothers milk

Group 2 – test article “inhaled”

African Queen

Purple Indica

Group 3 –“Other”

One Way ANOVA performed on the following test article groups:


Sativa (261 – 61%)

Other (105 – 24%)

Indica ( 65 – 15%)

Multivariate Tests performed for type of Cannabis, diagnosis, and change in variable


Pillai’s Trace

Wilks’ Lambda, and

Tests of Between-Subjects Effects

One Way ANOVA, Bonferroni, Post-Hoc tests performed for definition of diagnosis and treatment effectiveness

All tests performed using SPSS (Statistical Program for Social Scientists) Version 9.0

Question One
Are there physical, mood and perception changes resulting from use of the test article?

Table 5
Paired Samples t test

Comparing means before and after

95% confidence interval (2-tailed)








-3.72 +/-  3




1.92  +/-  3




3.02  +/-  4




-4.28  +/- 3




3.94  +/-  4




1.24  +/-  3

All are significant

Question Two
Does change in variable vary by method of treatment - ingested, inhaled or other?

Table 6
Question Two - Means of Variable Changes by Mode of Consumption

























































Question Two

Examination of the mean change (One way Anova – 95% confidence interval)

Significance was found for the following variables

Spasm p = 0.008

Insomnia p = 0.036

Table 7

Interpretation of ANOVA Method of Test Article Delivery

  • Group 1 is different than group 3
  • Average group 1 (ingested) = -4.39
  • Average group 2 (inhaled) = -4.50
  • Average group 3 (other) = -2.20
  • There is greater improvement in nausea (0.36) with ingestables vs. “other”
  • Ingestables and inhaled groups are not different

Question Three
Are changes in variables related to the different types of cannabis and primary diagnoses?

Table 8

Mean Change of Variables in Treatment Test Article Groups
























































One Way Anova – 95% CI


Table 9
Interpretation of ANOVA Method of Test Article Treatment Group

The Indica Group is different than Sativa Group


Average Indica = 3.06

Average Sativa = 1.53

Average Other = 2.22

There is greater improvement in energy (0.012) with Indica vs Sativa and “Other”

Sativa and Other treatment groups are not different

Interpretation of ANOVA Treatment Group

Indica was more effective to increase energy and appetite in any primary diagnosis group.

Use of any test article was effective in treating Nausea in the Orthopedic and HIV/AIDS diagnosis group.

Table 10

Mean Change in Variable by Primary Diagnosis





















































































One Way Anova 95% CI


**Small sample size unable to correlate

Table 11

Interpretation of ANOVA Method for Primary Diagnostic Group

The Orthopedic and Neurologic group are different than the “Other” primary diagnostic group.

There is greater improvement in Mood (p = 0.008) for the Orthopedic group vs. “Other”

There is greater improvement in Mood (p = 0.001) for the Neurologic group vs. “Other”


Average Orthopedic                 4.36                    

Average Neurologic                  4.04

Average HIV/AIDS                   2.87

Average “Other”                       1.33

Average Cancer                       2.64     

There is no difference between the AID/HIV and Cancer groups

Table 12
Interpretation of ANOVA Method for Primary Diagnostic Group

The Orthopedic group is different than the “Other” primary diagnostic group.

There is greater improvement in Energy (p = 0.43) for the Orthopedic group than “Other”



Average Orthopedic                 3.54                                 

Average Neurologic                  1.33

Average HIV/AIDS                   2.31

Average “Other”                       1.07

Average Cancer                       1.23     

There is no difference between the Neurologic, AID/HIV, and Cancer groups

Table 13
Interpretation of ANOVA Method for Primary Diagnostic Group

The HIV/AIDS group is different than the Orthopedic primary diagnostic group

There is greater improvement in Nausea (p =0.04) in the HIV/AIDS group than Orthopedic primary diagnostic group


Average Orthopedic                 -1.58                   

Average Neurologic                  -4.21

Average HIV/AIDS                   -4.54

Average “Other”                       -3.97

Average Cancer                       -4.18

There is no difference between the Neurologic, Other, and Cancer groups

Table 14

Interpretation of ANOVA Method for Primary Diagnostic Group

There is improvement in Appetite (0.010) for all diagnostic groups

There is no difference in mean change for the Appetite variable for specific primary diagnostic groups


Average Orthopedic                 4.57        

Average Neurologic                  3.50

Average HIV/AIDS                   4.44

Average “Other”                       3.08

Average Cancer                       3.00

  Table 15

Interpretation of ANOVA Method for Primary Diagnostic Group

There is improvement in Insomnia (p = 0.000) for all diagnostic groups

There is no difference in mean change for the Insomnia variable for specific primary diagnostic groups

Average Orthopedic                 -4.68       

Average Neurologic                  -4.66

Average HIV/AIDS                   -3.49

Average “Other”                       -2.93

Average Cancer                       -5.08


Table 16

Interpretation of ANOVA Method for Primary Diagnostic Group

There is improvement in Awareness (p = 0.000) for all diagnostic groups

There is no difference in mean change for Awareness specific to primary diagnostic groups

Average Orthopedic                 2.21  

Average Neurologic                  1.07

Average HIV/AIDS                   1.15

Average “Other”                       0.65

Average Cancer                       2.25

Correlation Analysis Question Four

Is change in mood correlated to change in energy?

p = .035*

Is change in mood correlated to change in pain?

p = .817

Is change in mood correlated to change in nausea?

p = .434

Is change in mood correlated to change in insomnia?

P = .647

Is change in mood correlated to change in awareness?

P = .073



There were observed changes in pain, energy, nausea, appetite, and awareness variables from the use of the test article.


ukcia banner
Use this banner to link to UKCIA

Page designed and maintained by UKCIA