DENNIS, Justice.

In this workers' compensation case, the injured worker suffered lumbosacral strain as a result of an on-the-job accident. Both the trial court and court of appeal awarded the plaintiff total temporary disability benefits calculated on the formula applicable for those employed on a "unit, piecework, commission or other basis." La.R.S. 23:1021(7)(d). We believe that the lower courts erred in determining the extent of the plaintiff's disability and the method for calculating his benefits. Accordingly, we must reverse with respect to these errors.

Allor, age 31, worked as a wire drawer for Belden Corporation and had been employed by that company approximately one year prior to his accident on March 16, 1978. He had worked 15 years before the accident doing strenuous manual labor without any physical problems. Allor became stiff and sore at work the day before the accident, but his low back pain became intense while he was lifting heavy spools of wire on March 16. He notified his supervisor who made an appointment with Dr. Sirikul, a local general practitioner. Dr. Sirikul examined Allor on March 20 and March 23. He found the plaintiff suffered from muscle spasms and tenderness in the lumbosacral area. The doctor took X-rays of the claimant's back which showed nothing abnormal. Dr. Sirikul diagnosed acute lumbosacral strain and referred Allor to Dr. Gamburg, an orthopedic surgeon in Alexandria. Dr. Sirikul stated that he believed that Allor's accident at work caused his back problems.

Dr. Gamburg treated Allor for several months and examined him numerous times. He confirmed Dr. Sirikul's diagnosis of lumbosacral strain. After three examinations, Dr. Gamburg found that the symptoms of Allor's back condition, i. e., pain and restricted mobility, had disappeared, and he released him to work in May, 1978. Upon returning to work, Allor again experienced severe localized lower back pain and muscle spasms and had to be hospitalized. Dr. Gamburg and a consulting neurologist found no disc herniation or neurologist abnormalities. Dr. Gamburg diagnosed the claimant's condition as recurrent lumbosacral strain. The doctor testified that he thought Allor could again return to work as of July 19, 1978 on a trial basis although he admitted that Allor's symptoms could reappear if the claimant returned to work.

Allor tried a second time to resume work several weeks later but found that the pain recurred and was too severe to allow him to continue. He sought additional medical treatment, but because Dr. Gamburg had moved his practice, Allor was thereafter treated by Dr. Weiss, also an orthopedic surgeon. Dr. Weiss first saw Allor in August, 1978 and continued periodic examinations through March, 1979. He also confirmed earlier diagnoses of lumbosacral strain and prescribed a back brace for Allor to wear which seemed to help the claimant. It was his opinion that Allor should not attempt to return to his job as a wire drawer or to any other heavy manual labor. Although Dr. Weiss thought Allor's lumbosacral strain would heal sufficiently by July, 1979 to allow him to be symptom-free and to pursue lighter work, the doctor predicted that Allor's lumbosacral injury would most likely recur if he tried heavy manual work.

The virtually uncontradicted evidence of the attending physicians and of the lay witnesses prove that, following the work accident, the employee suffered painful symptoms typically related to a low back injury which will recur upon his attempt to return to his former occupation and which prevent his undertaking other heavy physical labor. The lower courts ruled that this evidence supported an award of benefits to the plaintiff only during the temporary period in which he manifested physical symptoms resulting from the work accident. The primary question before us is whether a causal connection between the work accident and the plaintiff's permanent inability to return to heavy manual work has been shown by the plaintiff.

CAUSATION

The employee in a workers' compensation proceeding has the burden of establishing the disability and causal relation with the employment accident by a preponderance of the evidence. Nevertheless, it is not necessary for the experts to determine the exact cause of the disability in order for the employee to recover. The complaint need show only by a preponderance of the evidence that somehow the work accident caused the disability. Lucas v. Ins. Co. of North America, 342 So.2d 591, 595 (La. 1977); Malone Johnson, Workers' Compensation, § 252, p. 547 (2d ed. 1980).

On the specific issue of the causal connection between the accident and the disability, the claimant is entitled to the benefit of a presumption in certain cases. If an otherwise healthy worker suffers an accident at work and is thereafter disabled, it is presumed that there is a causal connection between the two, so long as the medical evidence establishes a reasonable possibility of such a connection. See Cadiere v. West Gibson Products Co., Inc., 364 So.2d 998 (La. 1978); Haughton v. Fireman's Fund American Ins. Cos., 355 So.2d 927 (La. 1978); Lucas v. Ins. Co. of North America, 342 So.2d 591 (La. 1977); Bertrand v. Coal Operators Cas. Co., 253 La. 1115, 221 So.2d 816 (1969); Malone Johnson, supra, § 259, p. 572-3. This presumption is not conclusive, but it forces the defendant to come forward with sufficient contrary evidence to rebut it.

Also we note that an injured worker is still entitled to compensation even though the work accident would not have been disabling to the average healthy employee. Numerous disability awards to injured employees have been granted when the evidence showed that some preexisting deficiency contributed to the disabling consequences of a work accident, even though the disease or condition might have brought about disability in any event in its ordinary course of progress. E. g., Roberson v. Liberty Mut. Ins. Co., 316 So.2d 22 (La.App. 3d Cir. 1975); Jones v. Douglas Public Service, 264 So.2d 267 (La.App. 4th Cir. 1972); Lum v. Employers Mut. Liability Ins. Co. of Wisconsin, 216 So.2d 889 (La.App. 2d Cir. 1968).

The defendant relies heavily on segments of Dr. Weiss's testimony to rebut the plaintiff's prima facie showing of a causal connection between the accident and the disability. The doctor refused to state that an accident precipitated the employee's inability to return to heavy manual work because he had not been given a "history of a specific accident on the job." He testified that he preferred to say that engaging in labor too heavy for a slight body had caused Allor's disability. Dr. Weiss also stated at one point that Allor would not suffer "any permanent impairment" from his injury.

When read in the context of his whole testimony and the entire record, however, these passages of Dr. Weiss's testimony do not constitute sufficient contrary evidence to rebut the presumption that there was a causal connection between Allor's work accident and his disability. First, it is clear Dr. Weiss only meant that Allor would suffer no permanent impairment insofar as his ability to perform light physical work was concerned. He testified repeatedly that Allor should not return to his job as a wire drawer or other heavy manual labor because this would reactivate his lumbosacral injury. Second, it is understandable that Dr. Weiss was reluctant to relate Allor's disability to an accident since the doctor did not have a "history of a specific accident on the job," although it is hard to understand why he was without a history of an accident which was well documented by the other attending physicians. Third, when he was asked whether he would relate Allor's "continued complaints" to a specific incident of heavy lifting, assuming that it occurred, the doctor said, "I think it is contributory [in] the sense that we know he got it, and that he will probably get it again if he goes back to doing heavy work."

The thrust of Dr. Weiss's testimony seems to be that Allor's physical disability was inevitable given the claimant's physique and work history. However, this does not prevent the claimant's accident from being a legal cause of his disability. We have said, "[i]t is immaterial that the disability could have been brought on by causes other than a work-related trauma, if, in fact, trauma on the job which meets the standards of accidental injury is a disabling factor." Parks v. Insurance Company of North America, 340 So.2d 276, 281 (La. 1976). See Bertrand v. Coal Operators Cas. Co., 253 La. 1115, 221 So.2d 816 (1969). We have held, moreover, that an accident which aggravates or accelerates a pre-existing condition is compensable even where disability is not caused by a single or specific incident. Parks v. Insurance Company of North America, supra, 340 So.2d 276; Chism v. Kaiser Aluminum Chemical Corp., 332 So.2d 784 (La. 1976).

In the present cast the timing of the proven disability can not be explained in any reasonable manner other than that the accident caused the disability. It is implausible to suggest that the disability would have occurred in the same manner and at the same time had the accident not occurred. Before his accident the plaintiff exhibited no disabling symptoms and adequately met the extreme physical demands of his employment. Allor passed his preemployment physical with Belden. The lay testimony establishes that the claimant was a good worker. Wage records indicate that Allor's output was consistently above that of the average wire drawer operator. After the accident and two attempts to return to work, the plaintiff is no longer able to perform his old job. Thus, we find not only did the claimant's accident cause him to be totally disabled for the temporary period in which he experienced substantial pain, but also caused his permanent inability to engage in heavy manual work without recurrence of these disabling symptoms.

EXTENT OF DISABILITY

We next reach the question of the extent of the plaintiff's disability. We find that Mr. Allor is now permanently and partially disabled within the meaning of La.R.S. 23:1221(3). We find that the preponderance of the evidence shows that Mr. Allor can no longer perform the "duties in which he was customarily engaged when injured or duties of the same or similar character, nature, or description for which he was fitted by education, training, and experience."

The medical testimony established that resumption of heavy manual labor would be detrimental to Mr. Allor's health and likely cause recurrence of his back problems. Indeed, Mr. Allor tried twice unsuccessfully to resume work. Dr. Weiss advised against the claimant resuming work which required heavy lifting, repeated bending or stooping, or standing for prolonged periods. Thus, any same or similar employment would be more physically demanding than the plaintiff could satisfactorily perform. Dr. Weiss did not rule out a wide range of jobs with lesser physical requirements however. It was his opinion that Mr. Allor could perform these type jobs without substantial pain or detriment to his health.

Mr. Allor completed ten and one-half years of school and obtained a high school equivalency degree. He has a varied work history having operated heavy equipment, driven trucks and wreckers, tended bar, and managed a service station. He was trained in the Marine Corps as a heavy equipment engineering mechanic. Also at the time of his accident, he was a cook in the Louisiana National Guard. He resigned his position after the accident during his period of severe pain. We do not find that the claimant has established a prima facie case of odd-lot status so as to be entitled to total, permanent disability benefits. His education, training, and physical condition are such that it cannot be said he is at a severe disadvantage in competing with others in any recognized labor market. See Oster v. Wetzel Printing, Inc., 390 So.2d 1318 (La. 1980). We believe less strenuous jobs could be performed by the plaintiff without detriment to his health or incurring substantial pain. We agree with the lower courts that Mr. Allor was totally disabled until June 27, 1979 and that this total disability was only temporary.

COMPUTATION FORMULA FOR AVERAGE WEEKLY RATE

Both lower courts ruled that Mr. Allor was employed on a "unit, piecework, commission or other basis," and thus, his average weekly wage should be calculated by resort to the formula provided in La.R.S. 23:1021(7)(d). It is undisputed that using the first formula, Mr. Allor is entitled to $99.28 weekly, whereas, if he is deemed an hourly worker, his weekly benefit amount would be $130, the maximum allowed.

While employed at Belden as a wire drawer, Mr. Allor received a base hourly rate of $3.325. He received overtime pay for any hours worked over forty each week at one and one-half his base hourly rate. In addition to the above, Mr. Allor got a "production percentage" depending on the amount of wire he produced. This incentive bonus was a varying percentage of his base hourly rate. The court of appeal found that in the four-week period immediately preceding his accident, the plaintiff's production percentage averaged 40%. Allor v. Belden Corp., 382 So.2d 206, 209 (La.App. 3d Cir. 1980). This can be equated to $1.33 more an hour (40% X $3.325) thus making Mr. Allor's hourly rate $4.655. Id. Despite the relative ease of making the above computation, the court of appeal, nevertheless, found that Section 1021(7)(a) "clearly contemplates a fixed, definite and unvariable rate of pay." Allor v. Belden, supra, 382 So.2d at 211. Overtime pay must be figured in the Section 1201(7)(a) calculation and often varies from week to week causing at least as complicated calculations as the production percentage involved here. See, e. g., Skinner v. Boise Southern Co., 364 So.2d 223 (La.App. 3d Cir. 1978). The court of appeal cited no authority for its contention that to be deemed an hourly worker one must receive a fixed and unvariable rate of pay, nor do we find any. We think the varying calculations for determining a worker's average weekly wage reflect a legislative attempt to protect adequately and equitably the earning capacity of injured workers who are compensated in differing ways. We find no intention by the legislature to make Section 1021(7)(d) a favored residual category so that all workers not precisely fitting any other category should be herded into it, indeed, the opposite seems true. "[Sub-section (d)] was apparently intended for situations in which the employee is paid solely on a unit or commission basis." Malone Johnson, Workers Compensation § 323 in 14 Louisiana Civil Law Treatise 91 (1980). [emphasis added] Professors Malone and Johnson further state that the multiplier of four in sub-section (d) was a concession to the logging industry in which many employees are paid on a per unit basis. Id. pp. 91, 100-101. It would be inequitable to deny Mr. Allor the full benefit of his production percentage which averages 40% of his hourly rate, nor would it clearly reflect his earning capacity. Thus, Mr. Allor's benefits should be computed under 1021(7)(a) since he is paid on an hourly basis. It is conceded that this will result in compensation payments of $130 per week minus wages Mr. Allor earns in any week after June 27, 1979 in accordance with La.R.S. 23:1221(3).

In summary, we conclude that the plaintiff is entitled to total, temporary disability compensation from the date of his accident, March 16, 1978, until June 27, 1979 at the rate of $130.00 per week. After June 27, 1979, the plaintiff is entitled to permanent, partial disability compensation in the amount of $130.00 per week minus any wages earned by him during the week. We find no error in the court of appeal's award of penalties, attorney's fees, medical, and travel expenses and thus affirm this award. We order that an additional $500.00 in attorney's fees be assessed against the defendant to cover the expenses incurred in prosecuting this appeal.

COURT OF APPEAL REVERSED IN PART, AFFIRMED IN PART, AND AMENDED.