Significant memory usage with COLUMN_GET

I've noticed significant memory usage with the repeated use of COLUMN_GET() in a query.

The structure of the table is:

CREATE TABLE mystats (

  id int(10) UNSIGNED NOT NULL,

  stats blob NOT NULL,

  PRIMARY KEY (id)

)

ENGINE = INNODB,

CHARACTER SET utf8,

COLLATE utf8_general_ci,

ROW_FORMAT=DYNAMIC;

This table is about 44GB (with 25 million rows). For each row, the blob contains on average about 120 dynamic columns (out of a set of 317 possible elements).
The datatype of these columns is primarily INT or DOUBLE.

The query tested performs limited select and selects a variable number columns from the stats blob.
The result set of this query is 209076 rows. The only difference between the queries in this test are the number of COLUMN_GET clauses used.
I recorded the memory usage of the same query using 3, 10, 30, 60, 100, 150, 200, 317 COLUMN_GET clauses in the SELECT statement.

I'm trying to understand why COLUMN_GET uses so much memory. The query with 3 COLUMN_GET clauses uses about 40MB of memory, but with 317 clauses memory usage balloons up to near 4.4GB.
Using COLUMN_JSON (replacing the COLUMN_GET clauses) doesn't seem to use any noticeable amount of memory running with the same query.

The form of the query is:

SELECT COLUMN_GET(stats, <stat> as DOUBLE), ... FROM mystats LIMIT 209076;

(where <stat> is the stat name as a string).

I'm not sure if this is a bug or expected behavior, but it appears anomalous to me.

Attached is a quick plot of memory usage of the mysql process as these queries have been executed. The y-axis is Resident memory usage in MB.
The values at each local maxima are

Elapsed time

7.537    10162.422  (3 clauses)

14.871    10272.504  (10 clauses)

23.611    10611.875  (30 clauses)

32.176    11129.355  (60 clauses)

42.730    11622.684  (100 clauses)

54.795    12225.711  (150 clauses)

70.572    12843.094  (200 clauses)

93.901    14540.383  (317 clauses)

The idle periods average about 10148.24 MB